tracker.js:10 Error: rawr i'm a dinosaur
at ChildProcess.child.on (/home/greg/resin/prog/etcher/lib/src/child-writer/index.js:92)
at emitTwo (events.js:106)
at ChildProcess.emit (events.js:191)
at handleMessage (internal/child_process.js:718)
at Pipe.channel.onread (internal/child_process.js:444)
a.(anonymous function) @ tracker.js:10
(anonymous function) @ /home/greg/resin/prog/etcher/node_modules/angular/angular.js:13708
logException @ /home/greg/resin/prog/etcher/lib/gui/modules/analytics.js:157
reportException @ /home/greg/resin/prog/etcher/lib/gui/modules/error.js:45
ImageWriterService.flash.then.catch @ /home/greg/resin/prog/etcher/lib/gui/pages/main/controllers/flash.js:82
processQueue @ /home/greg/resin/prog/etcher/node_modules/angular/angular.js:16170
(anonymous function) @ /home/greg/resin/prog/etcher/node_modules/angular/angular.js:16186
$eval @ /home/greg/resin/prog/etcher/node_modules/angular/angular.js:17444
$digest @ /home/greg/resin/prog/etcher/node_modules/angular/angular.js:17257
(anonymous function) @ /home/greg/resin/prog/etcher/node_modules/angular/angular.js:17483
completeOutstandingRequest @ /home/greg/resin/prog/etcher/node_modules/angular/angular.js:5955
(anonymous function) @ /home/greg/resin/prog/etcher/node_modules/angular/angular.js:6234
d.util.hasFunction.f.(anonymous function) @ tracker.js:3
Get this rawr i'm a dinosaurerror while writing the bz2 compressed version of the Raspbian Jessie image. Very slow writing (~3x slower than other compressed images), and breaks around ~95%.

Happens both with 1.0.0-beta15 AppImage, and git head npm start.
Haha wtf? This must be coming from the bz2 decompression module we use (can't think of anything else). Could be a funny way to say a package is deprecated?
Do you mind creating a new issue on that project, probably linking to the problematic image? The maintainer is very responsive and I honestly don't have the knowledge to debug such as bz2 algorithm issue.
Sure thing :) Filed bug: https://github.com/regular/unbzip2-stream/issues/10
Ha ha, I knew this would at some point show up in an application window. I did not write the actual decompression code (and neither did I author the error messages). I just wrapped it into a stream interface and packaged it. So, your guess is as good as mine, I'm afraid. The code is not very .. accessible unfortunately.
PR would be appreciated though!
No worries! I'll study how bzip2 works, and will try to fix or at least make sense out of it whenever I can allocate some time for it!
@jviotti @lurch actually seems to have figured it out already: https://github.com/regular/unbzip2-stream/issues/10
@jviotti where are we with this one?
We have https://github.com/cscott/compressjs, which is supposed to have
a fix for this issue, however they don't expose a standard NodeJS
implementation (they roll their own, which I can't get to play with
NodeJS, since its all done inside a single function), so we don't have a
good bzip2 decompression alternative yet.
On Thu, Dec 01, 2016 at 08:02:09PM -0800, Alexandros Marinos wrote:
@jviotti where are we with this one?
--
You are receiving this because you were mentioned.
Reply to this email directly or view it on GitHub:
https://github.com/resin-io/etcher/issues/734#issuecomment-264368271
--
Juan Cruz Viotti
Software Engineer
Maybe it's worth nudging the author of unbzip2-stream and asking him to pick up the bzip2 fixes from compressjs ? I could do that if you like?
@lurch That would be amazing! Thanks a lot!
@lurch Even better than nudging: a PR with test. I don't use that module in any of my projects and I am no expert on bzip at all, so I'm in no way more qualified to apply that patch. I will however merge a PR asap and republish the npm module.
Even better than nudging...
Haha, I'll see when I have time to put something together :smile_cat:
@jviotti how much work would be to either port the fix or use another library? This is our second oldest bug :)
I started hacking on it at some point, got too deep into the rabbit hole, and ended up abandoning it. I'm happy to give it another go.
I guess we could "won't fix" this. Not sure it's worth your time.
--
Alexandros Marinos
Founder & CEO, Resin.io
+1 206-637-5498
@alexandrosm
On Thu, Mar 2, 2017 at 8:18 AM, Juan Cruz Viotti notifications@github.com
wrote:
I started hacking on it at some point, got too deep into the rabbit hole,
and ended up abandoning it. I'm happy to give it another go.—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
https://github.com/resin-io/etcher/issues/734#issuecomment-283699639,
or mute the thread
https://github.com/notifications/unsubscribe-auth/ABLUCJjmnAqgk5P_1x198moV5rSDXgG4ks5rhuvAgaJpZM4KK2Ef
.
I guess there's no reason to not keep it open until a clearer way forward appears.
This might be fixed in [email protected] Let me know if it is!
Amazing stuff @regular , thanks a lot for this. I can't find any image that reproduces the issue, so I'll upgrade this on Etcher, and we'll see if it ever happens again, unless @imrehg still has the problematic image somewhere.
Unfortunately I no longer have a copy of @imrehg 's 2016-05-27-raspbian-jessie.img.bz2 file - I must have deleted it due to lack of disk space :-( (and @imrehg 's dropbox download link is no longer working)
So I instead attempted to recreate it by downloading, extracting and recompressing http://downloads.raspberrypi.org/raspbian/images/raspbian-2016-05-31/2016-05-27-raspbian-jessie.zip and that did then recreate the infamous dinosaur-error with [email protected] ;-)
And upgrading to [email protected] _did_ then eliminate the bug, so I can confirm v1.0.11 fixes this problem :smile:
Amazing news!!
Most helpful comment
This might be fixed in [email protected] Let me know if it is!