<Oksana>
jonwil: About localization patches: exception-en-US.diff should still be done, but the patch may need to be modified because of different line numbers (Mozilla changed their code slightly, but it does not affect -+diff, only the line numbers and pretext of the edit).
<Oksana>
bug137606.diff is no longer needed, since Mozilla-trunk does not mention File menu anymore.
<Oksana>
bug157237.diff should still be done. Not sure whether line numbers are still the same or changed. bug87973.firefox.hardcoded.diff should still be done. Bug 336029 for Mozilla, by the way.
xes has joined #neo900
<j4s0nmchr1st0s>
Oksana: Are you a programmer?
<DocScrutinizer05>
well, patch is extremely smart
<Oksana>
I did not make the patches. I just read them.
<DocScrutinizer05>
would you like to _read_ my little fancy at ~flashing? :-)
<dos1>
DocScrutinizer05: well, one of my ex-mozilla devices is still untouched, so I could reflash it
<DocScrutinizer05>
dos1: great!
<DocScrutinizer05>
it's made for exactly that
<Oksana>
Interesting script. Too complicated to read through, at least when I do not actually use it. :-)
<DocScrutinizer05>
complicated? it's only a 82 lines, some 20..30% of those comments or empty
<dos1>
RX-51_2009SE_21.2011.38-1_PR_COMBINED_MR0_ARM.bin is not known (has no checksum line in local file md5sums.txt for checking file integrity). Aborting.
<DocScrutinizer05>
ummmph
<DocScrutinizer05>
fixed
<DocScrutinizer05>
but it's easier to do `echo '095259c2380e894dc1d6a2999526ec9f RX-51_2009SE_21.2011.38-1_PR_COMBINED_MR0_ARM.bin' >>md5sums.txt
<Oksana>
What is libxul-sdk? Because there is some trouble around toolkit
<DocScrutinizer05>
I wonder if I should do sth about gaining root in the script
<dos1>
yep, I read the script before so I knew I should launch it as root
<DocScrutinizer05>
prticularly since wget shouldn't run as root, but flasher must run as root, also the blacklist hack and modprobe needs root permissions
<dos1>
~flashing doesn't say so though
<DocScrutinizer05>
:nod:
<dos1>
hehe
<dos1>
./flash-it-all.sh: line 65: 0xEEEE-3.5_2.5.2.2/EEEE-3.5: Permission denied
<DocScrutinizer05>
ugh
<DocScrutinizer05>
ffs
<DocScrutinizer05>
ls -l 0xEEEE-3.5_2.5.2.2/EEEE-3.5
<DocScrutinizer05>
or did you store the maemo-my-private-workdir/ on a nonexecute fs?
<DocScrutinizer05>
dos1: could you refresh please?
<Oksana>
jonwil: About crashes: 854ab30d101e.diff was already fixed by Mozilla, so no longer needed. GetChannelIsUnsafe_bug117540.diff should still be done. attachment.cgi%3fid=46259.diff was already fixed by Mozilla, so no longer needed.
<jonwil>
Put the information about the patches in the tmo thread otherwise it will go missing :)
<Oksana>
jonwil: Will try to.
<jonwil>
Will email romaxa with a nice well-written well-thought-out email I guess :)
<DocScrutinizer05>
GRRRR! could somebody please help me out how to store+quit in EMACS?
<DocScrutinizer05>
nm, done
<j4s0nmchr1st0s>
Oksana: Is there a prototype yet?
<Oksana>
j4s0nmchr1st0s:<DocScrutinizer05>a pity but what can we do? proto_V2 will not appear before end of year
<Oksana>
So, proto_V1 was done, some time ago, and proto_V3 will be the final prototype, equivalent to mass-production-device.
<DocScrutinizer05>
at least when we're lucky :-)
<DocScrutinizer05>
that's why we take a great deal of care to get everything right in proto_V2 already, so no surprises in proto_V3
<jonwil>
hmmm, dont know what to say in this email to romaxa (subject in particular)
<Oksana>
"MicroB: new Gecko" ?
<jonwil>
How about "Updaring Gecko in Microb"
<Oksana>
Updating*, even :-)
<Oksana>
Or "MicroB and latest Gecko"?
<DocScrutinizer05>
how about s/MicroB/fremantle browserd/
<DocScrutinizer05>
aiui microB is mainly the closed UI part
<Oksana>
All of this sounds alright. And MicroB probably sounds catchier, and more encompassing, than browserd. Not sure, though.
<DocScrutinizer05>
"updating MicroB's rendering engine from Gecko to <whatever> - help wanted"
<j4s0nmchr1st0s>
Oksana: Send me a prototype.
<Oksana>
MicroB-engine, for short.
<Oksana>
j4s0nmchr1st0s: Device prototype is not cheap. And I do not have one. Who is the head of Neo900 UG?
<jonwil>
"Updating Gecko in microb-engine to a newer version" works
<jonwil>
or better yet "Updating Gecko in microb-engine to a recent version"
<Oksana>
j4s0nmchr1st0s: If you want a prototype, pre-order one. Around 1500 euro, or what was it? proto_V2 will not appear before end of year
<DocScrutinizer05>
I'd not feel excited to rwad such mail when I had touched that stuff a 4 years ago last time. Subject sunds like a press announcement
<DocScrutinizer05>
read*
<jonwil>
how about "Help wanted updating Gecko in microb-engine to a recent version"?
<DocScrutinizer05>
"wow, that guy is updating microb-engine, whatever that means. Relevant for me? no. Next mail"
<jonwil>
I think "Help wanted updating microb-engine to a more recent Gecko version" works
<DocScrutinizer05>
you start to get an idea what it sounds like when werner, sebastian and me are drafting a new newsletter ;-)
<DocScrutinizer05>
we can do that game a full week
<Oksana>
Hopefully, he is still thinking about Linux-phone and Firefox-on-mobile-Linux.
<j4s0nmchr1st0s>
1500!
<j4s0nmchr1st0s>
eu
<Oksana>
Pre-order for final device is 100 euro.
<j4s0nmchr1st0s>
I used to get hardware for beta testing free.
<Oksana>
Estimated price for final device is 700 euro.
<Oksana>
Prototypes are hand-made and thus costly.
<j4s0nmchr1st0s>
Sign up for a reseller program Oksana
<Oksana>
Why reseller? Of prototypes, specifically?
<DocScrutinizer05>
j4s0nmchr1st0s: the project is completely financed by donations. Do you think it's fair to Oksana and all the other donors to use their money to build a proto for 1400EUR and send it to you for free?
<j4s0nmchr1st0s>
Depends on how many grants the company gets.
<Oksana>
If I am lucky, I will buy several final devices, and sell them to second-wave buyers. For the same price as I got them. Just for the sake of increasing the size of the batch.
<j4s0nmchr1st0s>
Let me look at the budget.
<DocScrutinizer05>
or let me put it this way: we already have ~300 betatesters who would *love* to get a prototype for a little less than the 1400 bucks it costs to build them
<Oksana>
Grants? No grants, just donors. Aka 'future owners of devices'.
<jonwil>
his last github contribution was to nemomobile-packages/gecko-dev on 26th september 2014
<jonwil>
so he is obviously active
<j4s0nmchr1st0s>
Is this a business or what that isn't a very well planned business model.
<j4s0nmchr1st0s>
Sounds more like you are building phones for James.
<DocScrutinizer05>
sounds like you're a really smart guy
<DocScrutinizer05>
you're welcome to donate a 500k EUR so we could easily hand out free prototypes to betatesters
<Oksana>
j4s0nmchr1st0s: It's going to be high-quality phone. With stereo speakers, stereo microphone, and such. With pressure-temperature sensors. With accelerometer and gyroscope. With resistive dual-touch. Read the specifications.
<j4s0nmchr1st0s>
Did you see that movie The Matrix?
<j4s0nmchr1st0s>
Compare the audio quality to the TP7
<Oksana>
With dual-SIM, NFC, infra-transceiver and FM transceiver.
<Oksana>
I mean, DragonBox-Pyra can optionally include cellular module
<Oksana>
"Wifi, Bluetooth and optional 3G/UMTS/GPS"
<DocScrutinizer05>
>>When some company releases a new device, it’s fairly easy to jump on your gun and point out “Hey, Are You Stupid? Why didn’t you choose THAT SoC instead? It’s more powerful!”. Well, most people actually forget one of the key market constraints: Supply.<<
<j4s0nmchr1st0s>
The phone also needs an FRS band for maintaining rugulatory network silence.
<j4s0nmchr1st0s>
If it cannot go local its just another consumer product.
<j4s0nmchr1st0s>
DocScrutinizer05: Stop what , chatting?
<DocScrutinizer05>
yes, this type of noisy meaningless chatter
<j4s0nmchr1st0s>
The suggestion about transmetta is not about crunching power but about compatibility.
<DocScrutinizer05>
IOW: read the links I poted above, read http://talk.maemo.org/showthread.php?t=91142*completely* from first to last post, then come back and ask your questions or even contribute with suggestions and criticism. But please not before you did all that
<Oksana>
j4s0nmchr1st0s: Or join ##chat : The problem with additional, country-specific features is that they add complexity (and cost), and are useless for many consumers
<j4s0nmchr1st0s>
But James.
<Oksana>
That's what hacker-bus is for.
<DocScrutinizer05>
Oksana: he won't know what's hackerbus, never even heard of. He doesn't know what CPU the device is using and why. Futile noise
<j4s0nmchr1st0s>
Same as the nook.
<j4s0nmchr1st0s>
Omap
<Oksana>
Maybe, he will get curious. Instead of trying to push his beloved features into the device, he will research how to add them to the device, without imposing costs onto people who would not want these features.
<j4s0nmchr1st0s>
My suggestion was for the same stated "why" compatibility.
<DocScrutinizer05>
sorry, I fail to parse you posts
<j4s0nmchr1st0s>
Amd knows exactly how to compete with intel.
<DocScrutinizer05>
for a rule of thumb: we are refining and discussing specs of this device with roundabout 300 peaople for over a year now. Guess how often we heard suggestions like yours already, and how often we mumbled "not *again* please" while typing an answer
<DocScrutinizer05>
and neither AMD nor intel is *any* relevant in this channel and for this device
<j4s0nmchr1st0s>
DocScrutinizer05: Do you know anything about transmetta?
<Oksana>
"Omap", "nook": what is that about? Compatibility - with what - can be achieved by using Transmeta? Neo900 strives for compatibility with Nokia N900, and thus uses TI processor, just a bit newer one.
<DocScrutinizer05>
:-/
<Oksana>
Transmeta seems to be used for laptops and workstations, not mobile phones.
<j4s0nmchr1st0s>
A phone needs not compete with intel but with other phones. The transmetta chip can do morphing.
<DocScrutinizer05>
and now please stop it, finally
<j4s0nmchr1st0s>
I like sugar.
<DocScrutinizer05>
when you're so convinced about your transmeta, I suggest you start your own phone building business and use your advantage over all the competitors, incl Neo900
<DocScrutinizer05>
and statements like your last one are no longer tolerated in this channel, I asked you two times already to reduce noise
<Oksana>
j4s0nmchr1st0s: Go into PM? DocScrutinizer does not enjoy noise
<j4s0nmchr1st0s>
Oksana: If you insist, go on, privmsg.
<jonwil>
Lets hope it results in something useful/positive
<jonwil>
He said "All you need is:" "1) Build latest GTK firefox (gecko engine) on Neo900 SDK" "2) rewrite microb-eal mozilla using latest Gecko engine API's.". Going to reply :)
<Oksana>
:-)
astr has quit [Ping timeout: 250 seconds]
astr has joined #neo900
Oxyd76 has joined #neo900
Oxyd76 has quit [Ping timeout: 250 seconds]
Oxyd76 has joined #neo900
j4s0nmchr1st0s has quit [Ping timeout: 244 seconds]
<fling>
Is there anything new about the project? Not much news on the site.
Sveta has joined #neo900
Svetlana has quit [Killed (kornbluth.freenode.net (Nickname regained by services))]
Sveta is now known as Svetlana
Sveta has joined #neo900
Sveta has quit [Remote host closed the connection]
Svetlana has quit [Quit: rebooting]
Svetlana has joined #neo900
jonwil has quit [Ping timeout: 255 seconds]
jonwil has joined #neo900
Oxyd76 has quit [Ping timeout: 250 seconds]
che1 has joined #neo900
jonwil has quit [Quit: ChatZilla 0.9.91 [SeaMonkey 2.30/20141013232806]]
jonwil has joined #neo900
Defiant is now known as Humpelstilzchen
<Oksana>
fling: Since the speakers of N900 were knowles TINY (not manufactures anymore), Neo900 is likely going to use knowles GRAND. They have slightly different dimensions, but *might* be feasible. http://neo900.org/stuff/joerg/random-media/speakers-IHF/
Oxyd76 has joined #neo900
<bencoh>
wow, so much dust
arcean has joined #neo900
<dos1>
DocScrutinizer05: Image(s) flashed successfully in 33.931 s (7543 kB/s)!
<DocScrutinizer05>
good
<kerio>
yay
sixwheeledbeast has quit [Ping timeout: 260 seconds]
<mvaenskae>
still the amound of replies is a nice sight for sore eyes, close to 200 messages and on my second page of subreddits i am subscribed to :)
<DocScrutinizer05>
and many of them would need a reply by somebody registered and entitled to post to reddit. Like the lord_skittles guy who obviously can't discern between open hardware and open source software
<DocScrutinizer05>
having a modem module that has whatever firmware on chip doesn't flaw the claim of "open hardware", just like using a ambient light sensor with a tiny MCU in it that does the A/D and calculations and I2C communication doesn't mean the device that's using this chip and clearly specifies how to talk to it via I2C is a closed hardware
<DocScrutinizer05>
heck, for most of the slightly more complex than "just 4 gates" chips you don't even know if they might have a MCU inside, and for sure you don't have access to any firmware running on such MCU
<DocScrutinizer05>
but all this is hardly related to *open hardware*
<jonwil>
treating such things as a black box is the right way
<jonwil>
i.e. you know what goes in and what goes out and you can be sure its not doing anything weird because its not possible for it to access things it would need to access to do weird things
<DocScrutinizer05>
in the end you need to do exactly this (treat as blackbox) with every single chip and even every single transistor you use
<DocScrutinizer05>
yep, exactly
<mvaenskae>
DocScrutinizer05: anyone can register on reddit btw, no need to be entitled from what i can tell
<DocScrutinizer05>
I for sure won't do that
<jonwil>
btw I am getting closer to tracking down a particular revision in mozilla hg that matches PR1.3 microb-engine
<DocScrutinizer05>
I probably once tried and it had hurdles of the type "only works on FireFox" or "you need google+ for that" or whatever
<jonwil>
the last relavent changelog entry for microb-engine was Sync up engine to last mozilla 1.9.2 branch and the last 1.9.2 branch as of that date is GECKO1922_20100315_RELBRANCH which is what I am now pulling
<DocScrutinizer05>
or simply "please fill that form which asks for a humble 182 personal dates of yours"
<jonwil>
a diff should tell me how close I am to matching things up
<jonwil>
I would say that microb-engine is up there as one of the most complex pieces of FOSS software on the entire N900
<DocScrutinizer05>
protestor, 19h ago >>There are concerns that even offline work can be compromised, since the baseband processor can run code on the other processor<< Simple but necessary answer: "NO Sir, not on Neo900"
<mvaenskae>
DocScrutinizer05: is that in regards to reddit? to post you need a reddit account consisiting of nick/password and if wanted an email :) (just to clear it up)
<DocScrutinizer05>
I'm generally not interested in registering at facebook, reddit, google+, twitter, younameit
<mvaenskae>
DocScrutinizer05: just wanted to clear it up :) it is by far the least worrysome of the large websites
<mvaenskae>
DocScrutinizer05: maybe it would be good for you to pick the wrong things and correct them, post them to the maemo forums and them dos posts a link to there that everyone can then see
<mvaenskae>
or dos can just copy-paste them directly into the responses
che1 has quit [Ping timeout: 260 seconds]
<DocScrutinizer05>
ratatask 2 points 2 hours ago >>People have already built their own GSM base station (http://www.openbts.org), and the baseband for one old motorola phone (http://bb.osmocom.org)<< S/He missed to notice the not so obvious link between osmocom and openmoko. Check Harald Welte, Dieter Spahr, among others. All are members of both the osmocom team and the openmoko crew
<mvaenskae>
DocScrutinizer05: out of interest, how legal was the base station?
<DocScrutinizer05>
100%
<mvaenskae>
:o
<DocScrutinizer05>
with a temporary licence
<mvaenskae>
ahhh
<DocScrutinizer05>
which is kind of funny since that licence afaik does _not_ require evaluation and certification of the protocols/"firmware" used, so such a test BTS could legally do rogue things (which it kinda did on CCC ;-D )
<DocScrutinizer05>
directhex 1 point 4 hours ago >> But it's not a maxed out iPhone 6. It's essentially an OMAP3 with performance found on high-end phones from 2010 ... So over $1000 for a phone from 2010 in terms of spec<< DUDE, go buy a antique car e.g. from the 1950s, ALL NEW and in mint condidtion. Then guess how much its value in terms of spec and how that compares to contemporary cars with same price
<mvaenskae>
DocScrutinizer05: 1950s? the neo900 ain't that old...
<DocScrutinizer05>
your point being?
<DocScrutinizer05>
>>I can't be the only person who thought that the phone was literally transparent?<< HUH what?
<bencoh>
no offense, but neo900 isnt exactly an "antique car" ;)
<mvaenskae>
the neo900 is more like a pontiac firebird
<bencoh>
there is no added value in it being "from the old days"
<bencoh>
mvaenskae: :]
<mvaenskae>
bencoh: there is, my parents bought a microwave over 20 years ago... it still works
<mvaenskae>
try that with new age stufff... nah, not gonna happen
<bencoh>
hmm ... point taken :)
<DocScrutinizer05>
it is. exactly for e.g. using a "old days" concept of not integrating modem into SoC, a concept that was essential for recent years' price drops on smartphones
<bencoh>
(though I'm not sure it's old enough for that)
<mvaenskae>
also my dad reapired the 20 year old washing machine regularily for the past 2 years just because he could
<bencoh>
DocScrutinizer05: yup, but we could go for a shiny-new-imx6 and have the same result
<mvaenskae>
would be stupid if just to change the coal one would need to buy a whole new engine
<bencoh>
DocScrutinizer05: yeah, I know about that, but it doesnt say why imx6 wouldnt be good enough for our purpose ;)
<bencoh>
(afaik it has the same issue than omap/powervr)
<DocScrutinizer05>
you should compare Neo900 to e.g. TI Zoom-II which sold for almost 2000 bucks iirc
<DocScrutinizer05>
rather than a silly insecure mainstream mass market smartphone like iPhone or Galaxy
<bencoh>
hmmm, okay, it does (logistical reason)
<DocScrutinizer05>
when even TI themselves cannot go cheaper for their development platform, how could we?
<DocScrutinizer05>
add to that that we want to be binary compatible to fremantle/N900 and there you are
<mvaenskae>
DocScrutinizer05: the iphone is pretty bold, not insecure; look at how well it advertises just how nice it looks!
<DocScrutinizer05>
I pondered for a long time whether to go for OMAP5 like pyra, but decided that it's probbaly not compatible enough and too high risk
* DocScrutinizer05
wonders what's wrong with his sense of humer. Notices that frowning is the regular reaction on such jokes, rather than smiling
<bencoh>
DocScrutinizer05: compatible enough with ? software ?
<DocScrutinizer05>
with N900
<bencoh>
yeah, but software-wise ?
<DocScrutinizer05>
err, what else? color wise?
<mvaenskae>
bencoh: look-wise
<bencoh>
:-))
<mvaenskae>
DocScrutinizer05: how much more R&D + unit cost (just the part) would the switch have caused?
<mvaenskae>
to omap5
<DocScrutinizer05>
for OMAP3 we have lots of existing apps and even several closed blobs proven to work in maemo
<bencoh>
I dunno, there might have been some hw bus you might want to use the same way ... :)
<mvaenskae>
DocScrutinizer05: is the omap5 not necessarily compatible to omap3 code?
<DocScrutinizer05>
mvaenskae: lemme put it this way: OMAP3 is a tested proven design, for OMAP5 we're wandering terra incognita
<DocScrutinizer05>
porting a partially closed OS to a new platform is hard enough already, even without changing the CPU during that enterprise
<mvaenskae>
but we want incognito mode per default D: (joking aside, i understand your concerns fully)
che1 has joined #neo900
<DocScrutinizer05>
csolisr 2 points 10 hours ago >>Laws say that the firmware must be compliant with the guidelines of regulation organisms (such as the FCC), not that the firmware must be proprietary. If a team develops a patent-free alternative and it's approved as compliant with said guidelines, no problem.<< DANG that person completely failed to grok what "cerification" means
jonwil has quit [Quit: ChatZilla 0.9.91 [SeaMonkey 2.30/20141013232806]]
<DocScrutinizer05>
>>However, the PowerVR drivers and baseband firmware are still closed-source. The idea their claim of "100% Free Software stack" is complete bullshit.<< that statement is complete BS, please RTFFAQ
<mvaenskae>
DocScrutinizer05: reading is a difficult task, comprehending even more so :)
<DocScrutinizer05>
could you please post that one line to this reddit thread, ideally in answering some of the BS talkers?
<mvaenskae>
DocScrutinizer05: i can post those links in a few likely (currently having lectures); please give me some sample direct links to the messages where i should reply to and i will post those there :)
<DocScrutinizer05>
>>What about evernote, amazon, ebay, aliexpress, humble bundle games, camscanner, cerberus, google drive, dropbox, car sharing apps, public transport apps, bank apps, flappy bird, ifttt, ingress, instagram, pushbullet, reddit, shazam, tripadvisor, waze. none of those was available when my n900 broke.<< this is really completely weird. How's for example reddit depending on an *APP* to use it?
<DocScrutinizer05>
""DUDE, go get an iPhone, so you may pay for reddit app""
<DocScrutinizer05>
mvaenskae: well, basically you can post it top level, it's obviously a mandatory read for all those who posted in that thread
<mvaenskae>
DocScrutinizer05: the first line of OP literally has /faq already even in bold
<DocScrutinizer05>
>>App store was painfully slow as well because it did a full reload of the repositories after installing one app from the GUI. Installing from the command line didn't have that problem.<< get speedyham! and be happy!
<mvaenskae>
i might need to break rediquette by posting a trojan that opens those links you posted every 2 seconds in a full screen window + screen reader at full volume
<DocScrutinizer05>
>>...the N9 has whatsapp and a music player.<< -- huh? and the N900 didn't, or what?
<mvaenskae>
the n900 never had a music player... it had a media player ;)
<DocScrutinizer05>
ooh, yes :-D
<DocScrutinizer05>
h3llm17 6 points 23 hours ago >>Resistive touch screen? What a joke.<<
<DocScrutinizer05>
Glimt 2 points 14 hours ago >>unless you use an active pointing device, like in Samsung's Galaxy Note line.<< -- which isn't working on capacitive principle at all
<mvaenskae>
DocScrutinizer05: so, should i still post the links? i mean it is part of the very first line of the OP
<DocScrutinizer05>
I thnk it cannot hurt
<mvaenskae>
ok, i will do that after changing my rooms :)
<mvaenskae>
give me some moments :)
<DocScrutinizer05>
>>maybe even modular so you can select then later upgrade a your camera module, put some other sensor there instead or have nothing.<< YEEEHAAAA FINALLLLY! phonebloks. No discussion about a new phone without some geek jumping in and shouting "MODULES, it needs MODULES!!"
<bencoh>
:D
kolp has joined #neo900
<bencoh>
DocScrutinizer05: I think you're hurting yourself :)
<DocScrutinizer05>
well, same dude later on in same post writes >>the best feature of course would be a modular aerial, that way you can choose to have a gert big sticking out one rather than the slim lined ones so popular today or even an external plug in roof aerial to get top notch reception even in a basement office or campervan.<<. No more to say to comment that
mvaenskae has quit [Ping timeout: 265 seconds]
<DocScrutinizer05>
>> Was I the only one confused when the picture of the phone showed a device not constructed from transparent plastic as if it was a gameboy from the mid 90s?? Then I realized what the title meant, and I became a bit less excited.<< HMMM, we might offer transparent cases, should be available like silver, gold, and white ones ;-) Honestly it seems there's an actual ambiguity in our website. dos1, could you tackle that maybe?
<bencoh>
ambiguity ?
modem has quit [Ping timeout: 250 seconds]
Oxyd76 has quit [Remote host closed the connection]
<pigeons>
unfortunately more people probably care about the transparency of their case than the rest of the system
paulk-collins has joined #neo900
<DocScrutinizer05>
indeed
<DocScrutinizer05>
but those will not pay the high price anyway
<DocScrutinizer05>
they want a cheap geeky possibly transparent plastics case phone, no matter if it works or not
<DocScrutinizer05>
just to show off
Oxyd76 has joined #neo900
Oxyd76 has quit [Remote host closed the connection]
Oxyd76 has joined #neo900
paulk-collins has quit [Remote host closed the connection]
che1 has quit [Ping timeout: 250 seconds]
modem has joined #neo900
modem has quit [Changing host]
modem has joined #neo900
che1 has joined #neo900
<DocScrutinizer05>
yeah, almost done with schematics review
Nokiabot has joined #neo900
<Nokiabot>
Somebody post that reddit link please :)
<Nokiabot>
Btw about that post on ridiculing resistive touch i think that guy should be pointed to "look what i drew on my phone thread ;)maybe a link on neo900 faq too
<Nokiabot>
Docscrutinizer05^^^
<DocScrutinizer05>
[2014-11-10 Mon 14:39:07] <DocScrutinizer05> h3llm17 6 points 23 hours ago >>Resistive touch screen? What a joke.<<
<DocScrutinizer05>
[2014-11-10 Mon 14:39:07] <DocScrutinizer05> Engival 27 points 22 hours ago >>Yeah, it looks horrible: https://www.youtube.com/watch?v=66RBfrBgL2E<< -- LOL, exactly right answer
<Nokiabot>
Ah that was the link sorry for making it reposting
Nokiabot has quit [Quit: used jmIrc]
<DocScrutinizer05>
LOL! http://www.fsf.org/news/gluglug-x60-laptop-now-certified-to-respect-your-freedom, I honestly doubt they got FOSS software for the kbd and system management / fan processor (the one you need to patch when you don't like the weird kbd layout regarding blue and control key). Nor for the harddisk controller, nor for any other of the obviously hidden_from_FSF little MCUs inside that thing
<Pali>
do you have open source code for EC? :D
<DocScrutinizer05>
EC?
<Pali>
embeded controller with lot of fun...
<DocScrutinizer05>
:nod:
<DocScrutinizer05>
thermal management aka fan control is a nightmare on those thinkpads under linux
<DocScrutinizer05>
thanks to that system management controller
<Pali>
that part which mix keyboard, mouse, PS/2 ports data for i8042... also that implement keyboard fn keys, and I bet that EC also controls fan, brightness and other parts
<Pali>
maybe sensors (temp, hdd protection) are connected to that EC too..
<DocScrutinizer05>
yes, correct
<Pali>
firmware in EC implements also i8042 multiplexer for supporting more mouse devices at one time (touchpad, trackpoint, external mouse)
<Pali>
and once firmware is bad you cannot do anything
<Pali>
also what to do with firmware in touchpads?
<Pali>
I bet that fsf does not provide any free firmware for it
<Pali>
and what with intel amt which is part of intel mainboard and is flashed by official bios flasher package?
<Pali>
they provide open source version of that intel amt firmware too?
<Pali>
and microcode for wifi card?
<Pali>
this fsf "Respect Your Freedom" is just fun
<Pali>
and intel txt and intel ethernet code? :D
<DocScrutinizer05>
ooh right, there's for sure a MCU on NIC
<DocScrutinizer05>
as they most likely won't hardwire all the low layer ethernet stuff
<DocScrutinizer05>
just think WOL
<Pali>
when you are flashing official windows "bios" exe package it provides all above pieces and flashing them (except ec and touchpad)
<Pali>
yes, wake-on-lan
<DocScrutinizer05>
RYF is a big hoax basically
<Pali>
so they probably provides only some of above pieces and other are untouched...
<Pali>
so yes, hoax
* DocScrutinizer05
idly wonders if a genuine VT52 terminal would qualify for RYF
<DocScrutinizer05>
or think about a hypothetical setup where the GLES "blob" runs off-board on some remote server and you send GLES commands to there and server returns binary values to write to the GPU registers. Is it entitled for RYF? Why not?
<Pali>
and SMM handler for x86 SMM mode?
<Pali>
(x86 system management mode)
<DocScrutinizer05>
think of a cellular modem that hypothetically needs to download own firmware OTA on each boot, but offers no way to load the firmware from local APE. Eligible for RYF?
<DocScrutinizer05>
I'd say it is, but obviously shouldn't
<Pali>
and can be any device with any cellular modem RYF?
<DocScrutinizer05>
FSF/RMS literally "thinking only inside the box" for RYF
<Pali>
mobile operator can send java apps and start them in SIM card without any user interaction
<Pali>
just only via cellular network
<Pali>
it is ok for RYF?
<DocScrutinizer05>
unless you remove SIM ;-P
<DocScrutinizer05>
so the phone is RYF but the SIM isn't
<DocScrutinizer05>
rotfl
<Pali>
modem without sim card is ok :D funny
<Pali>
is there open source firmware for sim cards?
<Pali>
probably not
<DocScrutinizer05>
how about a blackbox peripheral with clearly defined command interface based on a very Very VERY Long Instruction Set? Where e.g. the "reset device" instruction is some 580kB in size?
<DocScrutinizer05>
what's wrong with VLI ?
<Pali>
and processor microcode burned into SOC?
<DocScrutinizer05>
"other controllers need initialization sequences of up to 500 64bit commands, ours only needs one single 580kB command"
<DocScrutinizer05>
or think of FPGA and the amount of config data needed to get sent to them before they start working, none of those bytes/words is actually explained in detail how they work internally *on lowest level* inside the FPGA chip. When we specify 20 commands a 4kB length to get sent to a WLAN chip to initialize it, what's so fundamentally different in that, in the end?
<DocScrutinizer05>
sense to anybody, since the CPU running the code is not publicly documented.
<DocScrutinizer05>
the "it needs to be FOSS" part in "when some peripheral needs firmware uploaded, then that firmware needs to be FOSS" is also quite funny when you look closer. Any manufacturer could claim thet their hackers developed the firmware code in hex and the hex dump of the binary is already the best source code you could get for that particular firmware. And even when it was in assembler mnemonics, who says that those mnemonics do make any
<Pali>
stupid question, but can be binary hex firmware licensed as GPL? :D
<Pali>
e.g. you will release some binary file as GPL (same as you can release text file with .c suffix as GPL)
<Zero_Chaos>
Pali: gpl requires code
<Pali>
Zero_Chaos: binary file is code itself
<Pali>
what is different between file with .c suffix and file with .bin suffix?
<Zero_Chaos>
Pali: gpl requires source code, not just executable code
<Pali>
any C interpeter can execute file with .c suffix
<Pali>
so file with .c suffix is executable code too (for C interpeter)
<Zero_Chaos>
sigh
<Zero_Chaos>
Pali: I'm not hear to debate you, read the license yourself, it's very clear
<Pali>
but really somebody can design CPU (with its own micorcode) which can execute C file directly
<Pali>
Zero_Chaos: I read it
<DocScrutinizer05>
and files with .py suffix are for sure executables
<Pali>
and in my opinion some hexdump can be source code too
<Pali>
why not?
<DocScrutinizer05>
exactly my point
<Pali>
yes file with .py or .sh suffix is also sorce code
<Pali>
and also it is executable code
<Pali>
and both are accepted by CPU/VM which understand that language
<DocScrutinizer05>
something being "executable" doesn't mean it's no sourcecode as well
<Pali>
and x86 assembler code? there is 1:1 mapping between binary instructions and text equivalents
<DocScrutinizer05>
remove the .bin ending and replace it by .txt ;-P
<DocScrutinizer05>
voila, your sourcecode
<DocScrutinizer05>
prove the opposite if you can!
<Pali>
or is source code defined with chars if ( c >= 32 && c < 128 )
<DocScrutinizer05>
not afaik
<Zero_Chaos>
just because it's executable doesn't mean it isn't source code, I agree entirely. However, if it's pure binary it's almost certainly not source code.
<DocScrutinizer05>
sourcecode is not limited to ASCII
<DocScrutinizer05>
actually I did write code in hex for several years
<Zero_Chaos>
from gpl:
<Zero_Chaos>
The “Corresponding Source” for a work in object code form means all the source code needed to generate, install, and (for an executable work) run the object code and to modify the work, including scripts to control those activities.
<Zero_Chaos>
that seems pretty clear to me
<Zero_Chaos>
if you are writing it in binary, then that's the source
<Zero_Chaos>
if you are generating it, that it's
<Zero_Chaos>
isn't*
<DocScrutinizer05>
0xCD jumps to a subroutine, 0xC9 returns from subroutine (IIRC)
<DocScrutinizer05>
etc pp
<DocScrutinizer05>
I literally edited the stuff into a hex editor. So what's the sourcecode then?
<Zero_Chaos>
DocScrutinizer05: you edited it, or originally wrote it, in a hex editor?
<DocScrutinizer05>
both
<Zero_Chaos>
DocScrutinizer05: then that's the source
<DocScrutinizer05>
exactly
<Zero_Chaos>
DocScrutinizer05: if it's generated, that's not source, if you wrote it, that's source.
<DocScrutinizer05>
binary == source
<Zero_Chaos>
the license is pretty clear on this
<Zero_Chaos>
DocScrutinizer05: more like binary sometimes ~= source
<bencoh>
(what it doesnt say is how it can affect your sanity level :>)
<DocScrutinizer05>
in this particular case it is
<DocScrutinizer05>
bencoh: oh yeah, particularly calculation of jump distances, without a hex calculator at hand
<bencoh>
:))
<DocScrutinizer05>
anyway, when I put that binary blob under GPL then it's absolutely OK for firmware upload to a peripheral, even under FSF's RYF rules
<DocScrutinizer05>
I just proved that it's sourcecode, and the tools to process the sourcecode and generate a executable from it are commonly available
<DocScrutinizer05>
so FSF is pretty much out of luck when WLAN manufs put their binary blobs under GPL and declare it been written in hex editor directly to the file
<Pali>
Zero_Chaos: another question, that with Knuth's TEX? If I run web2c on TEX source code and then I modify it (because C is more readable then some WEB), then I cannot release C source code without TEX source code (in case TEX code is GPL)?
<dos1>
DocScrutinizer05: "Honestly it seems there's an actual ambiguity in our website." nope, we don't use "transparent" in our communication. the reddit post had "transparent" in its title
<DocScrutinizer05>
nope, no problem, since TEX already is publicly available
<Pali>
then we will have another project, but instead WEB language was used C and instead C was binary
<DocScrutinizer05>
dos1: AAAAH! many many thanks! :-)
<DocScrutinizer05>
dos1: I already started wondering which drug they mixed into my beer to make me agree on something so obviously unprecise and ambiguous
<DocScrutinizer05>
dos1: I however agree with the notion uttered last day that we need tpo make more clear that Neo900 does NOT develop or ship ANY particular software / firmware with the device
<DocScrutinizer05>
we *will* ship some testing software to check the hardware functionality, plus a BSP which is basically the same
mvaenskae has joined #neo900
<DocScrutinizer05>
everything beyond (e.g. a working SHR system installation) is a give away we add in for free, but we don't sell it and don't support or warrant it in any way
<DocScrutinizer05>
and the particular amount and type of such giveaway software shipped with device is subject to change any time without prior notice
<DocScrutinizer05>
think ASUS motherboard, or beagleboard
<DocScrutinizer05>
we are actually NOT selling a *phone* (product), we strictly sell some hardware only
<DocScrutinizer05>
it turns into a phone only when user installs according software of his liking to it
<DocScrutinizer05>
just like a beagleboard is no html-webserver
<DocScrutinizer05>
you can turn it into one by installing linux and apache
<DocScrutinizer05>
in that sense original early state of FSO was a BSP for GTA02
<DocScrutinizer05>
it came with a cheesy dialer and some tools
<DocScrutinizer05>
wasn't meant for end customers for daily usage
<DocScrutinizer05>
and of course it shipped FSO middleware. Which been it's main purpose
<DocScrutinizer05>
its*
<DocScrutinizer05>
I'd love to do exactly same for Neo900, with FSO, together with Mickey maybe :-)
<DocScrutinizer05>
though I guess you're competent and eager to implement such FSO based BSP on your own
<DocScrutinizer05>
maybe with a little help on kernel driver level, where needed (CRTOUCH, sensors, etc)
<DocScrutinizer05>
one of those posts on reddit was pretty brillliant to the point regarding all the "doesn't ship with hardware" topic
<DocScrutinizer05>
dos1: how about simply quoting the above post on neo900.org frontpage?
<DocScrutinizer05>
couldn't have put it better
wazrus has joined #neo900
<pigeons>
love the tangental argument on which laptop rms uses
<DocScrutinizer05>
hehe >> My brother is one of those N900 users, and I get the feeling you can pry the N900 from his cold, dead hands (he has a second spare one in case his current one dies). I used his N900 for a few weeks as my primary phone, and I found it slow, cumbersome, frustrating, heavy, and generally unpleasant to use - yet I still loved messing around with it, since it felt like using a real computer, instead of a locked-down, idiot-proof "
<pigeons>
well its mostly open source, dont know bout free
<ds2>
:)
<ds2>
you mean free like "enter a drawing now" and get a lifetime of spam? :D
<ds2>
but free don't matter
<ds2>
the UI is downright unpleasant to us
<ds2>
e
<ds2>
requires too many taps to do things. not as win8 but bad enough
<pigeons>
its annoying to worry about bionic libc compatability too
nox- has joined #neo900
<ds2>
I don't see that as too bad
<ds2>
but then I am used to dealing with libc of the day (uClibc/glibc/etc)
<ds2>
if anything bionic, IMO, is a good thing that should be used in more projects
<Wizzup>
not musl?
<Wizzup>
musl++
<bencoh>
is bionic actually usable for real posix stuff now ?
<bencoh>
I remember it having crippled pthread support
<pigeons>
it has pthreads, no pthread__atfork(), no cancellation, etc
jonwil has joined #neo900
sparetire_ has joined #neo900
mvaenskae has quit [Ping timeout: 244 seconds]
b1101 has quit [Quit: b1101]
<Pali>
bios is and was not usable for real suff
<Pali>
s/bios/bionic/
kolp has quit [Read error: Connection reset by peer]
<Pali>
bios (as int api) too :D
wazrus has quit [Quit: Lost terminal]
j4s0nmchr1st0s has joined #neo900
unclouded has quit [Ping timeout: 265 seconds]
unclouded has joined #neo900
norly has joined #neo900
<Oksana>
I am not interested in registering at FB, reddit, google+, twitter, whatever, generally. But my preferred method is flooding the system with irrelevant information, when I have time to bother with it.
<DocScrutinizer05>
I just tested a blue LED since we were interested in some properties of those, for another topic we just chew on. the results: 3V4 @ 10mA; 2V3 @ 0.02mA with already visible blue light. I guess for a red LED the current threshold for visibility is roundabout same, while voltage may be even lower. Neo900 probably will have quite a number of inexpensive simple diagnostic LEDs on power supply rails and signal lines, each consuming a max of
<DocScrutinizer05>
20µA
<Oksana>
Interesting. So no need to measure the currents and such when LEDs already tell you what is happening?
<DocScrutinizer05>
yep, that's the idea
<DocScrutinizer05>
"what the heck, why is this thing still sending an IRQ signal while it should already be shutdown?"
<Oksana>
:-)
<DocScrutinizer05>
"why does CPU take 5s to react on a RING signal sent from modem?"
<DocScrutinizer05>
"why is master system clock running all the time, despite CPU supposed to be in zero-clock?"
<DocScrutinizer05>
"Thanks, dear customer, for this bug report. Could you please tell us if the LED marked "E5" in figure 9 of your manual is showing dim light or not?"
<DocScrutinizer05>
even "please make a photograph of the device in a completely dark room without flashlight, and send this picture. Thanks!"
<Oksana>
:-) Great customer support.
<DocScrutinizer05>
LEDs and resistors are cheap, and small
<DocScrutinizer05>
and now we know they also don't necessarily consume worrying amounts of power
<DocScrutinizer05>
of course it's my task as EE and architect to place those LEDs in a way so they usually are off and only light up in the exception-case
<DocScrutinizer05>
for everything except power rails this shouldn't be a problem
<DocScrutinizer05>
seems for 3 bucks we could get a 50
<DocScrutinizer05>
not that I'm expecting to find use for 50 diagnostic LEDs in Neo900
<DocScrutinizer05>
so make that 25 per device for 1,50 bucks
<DocScrutinizer05>
we can kick them out any time, to make device 2 EUR cheaper, for mass production ;-)
<ds2>
do you want to annoy the PnP programmer by changing your BOM? ;)
<DocScrutinizer05>
the what?
<DocScrutinizer05>
pick&place?
<DocScrutinizer05>
no isse
<DocScrutinizer05>
issue
<ds2>
pick and place guy
norly has quit [Quit: Leaving.]
norly has joined #neo900
norly has quit [Client Quit]
<DocScrutinizer05>
I doubt there's much manual work involved, except for applying a offset tuning patch to new gerber from last run
<DocScrutinizer05>
unless you move components around on PCB, in which case you obviously need to readjust the placement
<DocScrutinizer05>
or at least check if it needs adjustment
<ds2>
it should as simple as uncommenting the part that stuffs it but...
<DocScrutinizer05>
it actually is even simpler, you set a flag for a variant in eagle
<DocScrutinizer05>
then generate new Gerber from that, then apply offset adjustment patch to that as mentioned above, and done you are
<DocScrutinizer05>
basically a press of two buttons
<ds2>
you doing it on the PCB level...
<ds2>
i was referring to just not stuffing it by asking the PnP operator to DNP the LED + resistor
<DocScrutinizer05>
can't be done since they are used in several positions, we probably don't use sequenced tape P&P
<DocScrutinizer05>
unless we run short of slots for components on the P&P machine
<DocScrutinizer05>
if that happens, we would need to make specially sequenced tape, also quite silly
<freemangordon>
jonwil:hi! Any progress on that big mysterious structure?
<DocScrutinizer05>
(silly) making special tape with empty positions that is
arcean has quit [Quit: Konversation terminated!]
<freemangordon>
jonwil: just curious, not that I am trying to put pressure on you :)
<freemangordon>
BTW there are 3 more functions and xprot will be ready
<DocScrutinizer05>
\o/
<freemangordon>
yeah, though those re like spaghetti
<freemangordon>
*are
<freemangordon>
but still doable
<Oksana>
Silly off-topic question: most dialogs have two, at most three buttons. Like, Yes-No-Cancel. Why cannot it be done with one software button and two-or-three mouse buttons? Left=Yes, Right=No, Middle=Cancel?
<freemangordon>
Oksana: ever seen fremantle?
<Oksana>
Yes...
<Oksana>
It does not have mouse?
<freemangordon>
well, then you should have know things can be done some other way
<freemangordon>
sure
<freemangordon>
well, not really
<freemangordon>
it can have mouse
<Oksana>
But it would be cumbersome for walk-use outdoors.
<bencoh>
:D
<Oksana>
And even with dual-touch you could imitate two mouse buttons, at most: one finger or two fingers.
<freemangordon>
anyway, there you have the needed buttons to take the actions and you click outside of the "dialog" for cancel. not exactly what you want, but IMO it is more sane. what if you need a context menu, what you should do in your UI, click-n-hole the right button?
<freemangordon>
*hold
<freemangordon>
it is obvious that I am old-fashioned, but n-touch is too complicated for me :)
<freemangordon>
I simply can't grok the multi-touch paradigma
<DocScrutinizer05>
I'd often prefer push&drag buttons, kinda like lockscreen slider, just with more than one dimension/direction
<freemangordon>
anyway, /me is going afk
* DocScrutinizer05
too
<Oksana>
Alright :-) I was just struggling with desktop. When mouse auto-jumps to default dialog button, and I want to click No instead of Yes, it makes sense to wish for one unified dialog button, with three different actions tied to three mouse buttons.
<Oksana>
Clicking outside of dialog for cancel makes sense, especially for one-window interface (such as Fremantle).
<DocScrutinizer05>
for desktop I'd consider it a *really* *really* useful concept when unsolicited popup requesters would get delayed (at least they focus grabbing) until you stop typing on kbd for at least 1500ms
<DocScrutinizer05>
of course that last value should be adjustable by user, just like speed of doubleclick
<Oksana>
That's not unsolicited popup. :-) That's Close with saving, Close without Saving, Cancel Closing dialog.
<Oksana>
And I have to go through repetitive motions (including this dialog) a lot of times.
<DocScrutinizer05>
I can't count how often I typed in an incorrect new password instead of keeping the already pre-populated old one, when kmail suddenly pops up a requester informing me about "authentication to POP3 server XYZ failed! Please enter correct user name and password!"
<DocScrutinizer05>
ever thought about using hotkeys for those "close, save, cancel" requesters?
<Oksana>
Hmm... Cannot it use some kind of tray-tooltip, with textfield included?
<Oksana>
Hotkeys? Like, arrow right, space?
<DocScrutinizer05>
like enter, escape, "alt-S"
<Oksana>
I mean, I have a program which persistently shows a tooltip from its tray. It's highly visible, yet it does not steal focus or stop you from typing anything else.
<DocScrutinizer05>
afk, cya
<Oksana>
If such tooltip included textfields with pre-populated usename and password, it would not steal focus.
<Oksana>
Visibility: in bright sun or in dark room?
<DocScrutinizer05>
in dimmed room of course
<DocScrutinizer05>
in bright sun not even LCD backlight at blasting full power of 9V6@36mA is visible
<Oksana>
:-)
<DocScrutinizer05>
a device with 25 of such LEDs shining all the time would deplete a standard BL-5J battery from full to empty in as short as >100 days
<DocScrutinizer05>
110 actually
<DocScrutinizer05>
hardly any issue regarding maximum standby time, eh?
<DocScrutinizer05>
~25*0.08
<infobot>
2
<DocScrutinizer05>
already two bucks for 25 LEDs
<DocScrutinizer05>
real estate (assuming the R's are as big as those LEDs):
<DocScrutinizer05>
~0.8 * 1.6 * 2
<infobot>
2.56
<DocScrutinizer05>
~2.56 * 25
<infobot>
64
<DocScrutinizer05>
8 x 8 mm area
<DocScrutinizer05>
ok, i cheated a little since you can't pack them without any gaps between