Topic for #qi-hardware is now Copyleft hardware - http://qi-hardware.com | hardware hackers join here to discuss Ben NanoNote, atben / atusb 802.15.4 wireless, and other community driven hw projects | public logging at http://en.qi-hardware.com/irclogs
emeb has joined #qi-hardware
methril has joined #qi-hardware
FrankBlues has joined #qi-hardware
FrankBlues has quit [Remote host closed the connection]
DocScrutinizer05 has quit [Disconnected by services]
DocScrutinizer05 has joined #qi-hardware
DocScrutinizer05 has quit [Client Quit]
DocScrutinizer05 has joined #qi-hardware
kristianpaul has quit [Ping timeout: 256 seconds]
kristianpaul has joined #qi-hardware
kristianpaul has quit [Changing host]
kristianpaul has joined #qi-hardware
emeb has quit [Ping timeout: 248 seconds]
FrankBlues has joined #qi-hardware
LunaVorax has quit [Ping timeout: 260 seconds]
<wolfspraul>
nice, I see that xilinx reintroduced smaller artix-7 devices
<wolfspraul>
namely the xc7a20, 35, 50 and 75
<wolfspraul>
I think as soon as I can get physical xc7a20 I will switch to that in fpgatools (and drop support for xc6slx9)
<wolfspraul>
depends on price. I think xc6slx9 is about 7 USD now, if the a20 is < 20 USD it's time to switch imho
<wolfspraul>
it might well be around 10 USD right from the start (once it's available)
<wolfspraul>
time to speedup in fpgatools and move all that I think could be xc6 specific to one modular place...
urandom_ has quit [Remote host closed the connection]
pcercuei has quit [Quit: Bye]
FrankBlues has quit [Remote host closed the connection]
DocScrutinizer05 has quit [Read error: Operation timed out]
DocScrutinizer05 has joined #qi-hardware
wolfspraul has quit [Ping timeout: 248 seconds]
wolfspraul has joined #qi-hardware
emeb has joined #qi-hardware
qwebirc53692 has joined #qi-hardware
qwebirc53692 is now known as Fred
Fred is now known as Freddy
Freddy is now known as FreddyFlin
<FreddyFlin>
lol
<FreddyFlin>
hello
FreddyFlin has quit [Quit: Page closed]
emeb has quit [Ping timeout: 255 seconds]
<wpwrak>
wolfspraul: dropping support for chips that are still in their prime sounds like a horrible idea
<wpwrak>
i'd rather view it as an opportunity to implement support for multiple architectures (which you'll want sooner or later) with very little testing overhead
kuribas has joined #qi-hardware
jekhor has joined #qi-hardware
LunaVorax has joined #qi-hardware
xiangfu has joined #qi-hardware
<wolfspraul>
I will absolutely drop xc6slx9 support the same day I start with xc7a???
kilae has joined #qi-hardware
<wolfspraul>
that means I will try to modularize the codes and leave the old stuff around, but if the 'carry forward' overhead gets too big, I need to remove it entirely (still in the src history of course)
<wolfspraul>
there is no way I will load any sort of "multiple architectures" stuff upon me
<wolfspraul>
not even the slightest idea of it
<wolfspraul>
give me 100 or so engineers and we talk about it again :-)
<wolfspraul>
I want the project to have real-world usage, even if it's just my own hobby boards
<wolfspraul>
that is still better than a meaningless 'infrastructure' or 'framework' imho
<wolfspraul>
if someone comes forward to support xc6 (at that point), that's another thing but I do not assume that
<wolfspraul>
this planned change (to xc7) will only come after xc7 is in "its prime" anyway, i.e. when I can get the xc7a20 on the street
<wolfspraul>
that will take a long time I think, late 2013 at the earliest
<wolfspraul>
let's see what happens until then - full power ahead with xc6slx9 anyway
<LunaVorax>
Good morning
<wolfspraul>
one example - the xc6 has a rather unusual word width of 16 bits, whereas I think many (all) virtex and I guess also xc7 have 32-bit word width
<wolfspraul>
I need to see what this means in practice, but if it's very big it may just not make sense to make the one engine to rule them all
<wolfspraul>
at least not at this point
rz2k has joined #qi-hardware
<wpwrak>
hmm, if you have this sort of variables, it ought to make sense to include them as variables also in your library. that is, if you want an "fpgalib" if you want an xc7lib or an xc6lib, and leave the abstraction to a higher layer, that wold be another option.
<wpwrak>
but it seems that, however you tackle this, as long as you do it right, the effort of having xc6 coexist with xc7 should be rather small
<wpwrak>
so planning to abandon xc6 seems radical and unnecessary
<wpwrak>
a bit like saying "when i'm older, i will move to another town. and then i'll burn down this one." you'd be scared if someone says that :)
<wolfspraul>
that's all theoretical, but I am dealing with 30k (and growing) loc
<wolfspraul>
so yes, let's just see
<wolfspraul>
I am excited that the smaller (and presumably cheaper) xc7 devices are back *on the roadmap*
<wolfspraul>
that's all right now
<wolfspraul>
I don't expect them to be available for many more months, time I have to get more and more out of the xc6slx9
<wolfspraul>
also keep in mind - most of that structure is completely newly conceived, and you cannot abstract the unknown
<wolfspraul>
so I need to start with case A, B, C, then rewrite A & B because it makes no more sense, etc.
<wolfspraul>
I think more or less we agree, but I may be a bit more radical on only keeping 100% working and tested codes
<wpwrak>
well, that's the domain of automated regression testing :)
<wpwrak>
things like compilers are incredible fertile areas for that
<wolfspraul>
fpgatools is essentially a big test routine
<wolfspraul>
I write the tests first, then implement features
<whitequark>
TDD!
<wpwrak>
kewl
<whitequark>
wpwrak: wtf "kewl" means? urbandictionary is almost useless, presenting two opposite meanings
<wpwrak>
;-)))
<wpwrak>
it's a cooler form of saying "cool". in this case, without irony. (although i often use it in an ironic way. but then i tend to use irony a lot in general :)
<whitequark>
there really ought to be a form of markup for irony
<whitequark>
a russian reddit clone has more or less adopted red italic text, but I've yet to see this convention elsewhere
<wpwrak>
it's part of passing the turing test ;-)
<larsc>
<irony>Well, there is</irony>
<larsc>
;)
<whitequark>
wpwrak: haha, this is an excellent idea: make a CAPTCHA out of this!
<whitequark>
larsc: I'm not sure if using XML to demonstrate irony is ironic or not
<wpwrak>
larsc: you could use the <I>short form</I> as well
xiangfu has quit [Ping timeout: 240 seconds]
<whitequark>
wpwrak: our postgresql discussion is quite funny in retrospect
<whitequark>
I was able to design a system (for gathering web statistics) which is quite storage-efficient and also flexible, and all it uses is a few queries with aggregate and window functions
<whitequark>
I swear that someone... younger than me would attempt to use a fancy NoSQL database and map/reduce
<wpwrak>
i'd probably not use SQL either. people often greatly underestimate just how fast you can read and process even huge files these days :)
<wpwrak>
so this means that queries are usually cheap and you need a "real" database only when you have complex update patterns
<whitequark>
wpwrak: that doesn't make any sense in context, because the storage should be accessed over network and it also relates to data already in our SQL RDBMS
<whitequark>
(besides needlessly reinventing the wheel)
<whitequark>
and it's not that queries of a typical web application are that complex either (which is the sole reason of attempts at using nosql for storage in webapps looking successful at the first glance); SQL databases are simply known to be reliable and well-tested, if too complex for most applications.
<wpwrak>
well, your web server produces logs in linear files. so if all you need are queries for statistics or management purposes, a RDBMS is a bit overkill
<whitequark>
wpwrak: NO.
<whitequark>
we are currently migrating from exactly the system you've described
<whitequark>
first, what I want is to log access to resources. I don't care about URLs, I care about dissecting nginx logs with regexps even less, as it's slow and unmaintainable
<whitequark>
and of course after dissecting those logs, the data went to an RDBMS, because no, you do not want to query literally tens of gigabytes of logs to get monthly statistics
<whitequark>
what's worse is that this setup does not scale: at the point where you have two backends, all the hell breaks loose
<wpwrak>
okay, at tens of GB you need to aggregate in one way or another
<whitequark>
it's tens of GB right now. we don't even have a moderate amount of customers.
<whitequark>
you underestimate how much junk there is in webserver logs; and there are things you simply cannot do with those
<whitequark>
like quickly rewind to a particular amount of time
<whitequark>
er, point
<wpwrak>
lseek and binary search ?
<whitequark>
a poor man's index ;)
<whitequark>
that still does not solve network access
<whitequark>
no, NFS is not an answer
<wpwrak>
but so universal :) well, at least you could make it so if you wrote a few command-line tools. it's actually surprising that there are none that do that
* whitequark
shrugs
<whitequark>
because there are tools better suited for the job?
<wpwrak>
sort of a cat-range <sort(1)-options> [<from-key(s)>] [<to-key(s)>]
<viric>
systemd journal :)
<viric>
that's the trendy linux thing
<whitequark>
sqlite3. simple, compact, more reliable than any combination of unix tools combined.
<whitequark>
er, any amount. (I don't seem to be drunk...)
<whitequark>
wpwrak: I was always amazed of how they test sqlite3. 100% code coverage, where 100% means that they test every possible boolean combination in every conditional
<whitequark>
so for a && b || c that would be eight tests.
<wpwrak>
that only makes sense. why go for less if you've already done most of the work ?
<whitequark>
hm?
<wpwrak>
if you have a test for one combination it should be easy to test for the others too
<wpwrak>
besides, it looks nicer if you have a larger number of tests ;-)
xiangfu has joined #qi-hardware
<whitequark>
91 MLOC of them, to be precise
<whitequark>
for 81 KLOC of code
woakas has quit [Ping timeout: 248 seconds]
woakas has joined #qi-hardware
xiangfu has quit [Ping timeout: 252 seconds]
xiangfu has joined #qi-hardware
jekhor has quit [Ping timeout: 256 seconds]
kuribas has quit [Remote host closed the connection]
jekhor has joined #qi-hardware
xiangfu has quit [Ping timeout: 260 seconds]
xiangfu has joined #qi-hardware
jekhor has quit [Ping timeout: 276 seconds]
jekhor has joined #qi-hardware
kilae has quit [Quit: ChatZilla 0.9.90 [Firefox 18.0.2/20130201065344]]
jekhor has quit [Ping timeout: 256 seconds]
xiangfu has quit [Quit: leaving]
urandom_ has joined #qi-hardware
emeb has joined #qi-hardware
emeb has quit [Ping timeout: 248 seconds]
emeb has joined #qi-hardware
guanucoluis has joined #qi-hardware
emeb has quit [Quit: Leaving.]
Artlav has joined #qi-hardware
<Artlav>
Hello.
<Artlav>
Anyone know of an SSTV decoder for Nanonote?
jekhor has joined #qi-hardware
<Artlav>
Would it even have enough CPU power to do it in real time?
guanucoluis has quit [Quit: Leaving.]
<methril>
Hi all!
<whitequark>
Artlav: SSTV takes 3kHz of bandwidth
<whitequark>
you could even decode it on AVR, I guess
<Artlav>
So, not a fundamental problem. Anyone done it on nanonote?
jekhor has quit [Ping timeout: 264 seconds]
kilae has joined #qi-hardware
FrankBlues has joined #qi-hardware
kilae has quit [Quit: ChatZilla 0.9.90 [Firefox 18.0.2/20130201065344]]