Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
riedel
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
41 ms
·
691.
▲
by
riedel
4y ago
This is really difficult business and I recently ended in a dispute with my civil engineer wife because I wanted to move a shelf towards the center of the room. She is I guess right reminding me that trial and error is not a good strategy i
692.
▲
by
riedel
4y ago
The patent troll is what killed the business plan in the end because there was zero margin. The real problem seems to me that it is difficult to compete with a nicely integrated 30 usd echo dot if people rather pay it with their personal da
693.
▲
by
riedel
4y ago
I guess 'notable' means something like >= 0.01% in this case. But tech debt is an issue if nobody can effectively fork off anything if unhappy with the direction taken.
694.
▲
by
riedel
4y ago
I really wonder about press standards in this case. Both parties are not public figures. Is it normal 8n the US that both are named and pictures are published. I think here in Germany typically at least a pseudonym would be assigned to prot
695.
▲
by
riedel
4y ago
AFAIK Cyrus has JMAP support. We are using it for more than 20 years at work without any issues. However, we are lacking compelling client support for JMAP (via using SoGo via IMAP currently). Would be great to have a good static webclient
696.
▲
by
riedel
4y ago
Wonder why the art was "lost" (maybe between 2008 and now, as the title does not contain th word)? I guess it is true that with computer program one does not have to rely an books. However, looking at the civil engineering books o
697.
▲
by
riedel
4y ago
Also using Keepassxc. However, I really would love to have sth audited and with a documented threat model. I have the feeling that those programs are developed by enthusiasts and probably would not withstand a serious attack (but this not r
698.
▲
by
riedel
4y ago
In case anyone wonders about the resolution: https://www.dailymail.co.uk/news/article-2786723/London-skys...
699.
▲
by
riedel
4y ago
Might be good not only to show that YOLO and other do not fully generalize to the presented image domains but also that some specialized SOTA models e.g. for areal images perform well in the subcategories to actually calibrate the benchmark
700.
▲
by
riedel
4y ago
Funnily my wife bought just bought a stabilized monocular from kite optics (12x25 [1]) today . It really cool how to see how much more one can see with stabilization turned on. [1] https://www.kiteoptics.com/en/nature&#
701.
▲
by
riedel
4y ago
My point is that there is socioeconomic theories which were based on the goal of common ownership of all means of production (which in a digital society would often translate to code). I guess all movements that have tried to transition to
702.
▲
by
riedel
4y ago
Hopefully closing the mouth does not affect the postprocessing algorithm of the MRI software (remember tons of fMRI studies invalid due to software artifacts [1]). I guess it really far to easy to generate such results, but difficult to rea
703.
▲
by
riedel
4y ago
People have to understand that with many statistical processes increasing the average means also increasing the variance.
704.
▲
by
riedel
4y ago
Looks like worker [1] is an actively developed clone, that very much looks like the original (haven't tried it yet). [1] http://www.boomerangsworld.de/cms/worker/index.html
705.
▲
by
riedel
4y ago
Actually DOpus was the only dual pane Filemanager (maybe with the exception of Ghostcommander on android lately) I actually got used to and was really fond of. I still wonder how working with floppy disks (two drives and a 4MB ram extension
706.
▲
by
riedel
4y ago
a startup from our university [1] successfully builds on something like this for various automation tasks. [1] https://www.askui.com/
707.
▲
by
riedel
4y ago
Also working nicely both numeric and symbolic: https://smath.com (3MB windows binary or cloud version, russian freemium)
708.
▲
by
riedel
4y ago
We ditched our air compressor at work because there are strict work safety regulations (like regular maintenance). However, going off grid is probably a bigger risk in itself and realistic risk analysis is a hard task.
709.
▲
by
riedel
4y ago
Actually resume stuffing was the major reason (not to get a job but rather for positioning oneself). I wrote it in my free time actually (my income on the book is about a 10th of the time put into it). That it is free to the students is jus
710.
▲
by
riedel
4y ago
I wonder if you could actually rather quantify date related errors by analytics. I actually booked the wrong return flight once because the date picking logic was too intelligent for me. There should be many signals pointing to this.
711.
▲
by
riedel
4y ago
Can you quote the sentences your are concerned about. While one could maybe make some non-sequitur conclusions by reading quickly, I do not see that much sensational reporting. The question really is what 'large amounts' mean when
712.
▲
by
riedel
4y ago
I published a book with degruyter and there is no proofreading at all. No support to get e.g. the colours right for printing. They only give you some example books, a rough latex template and minor advice on content and audience. Once it is
713.
▲
by
riedel
4y ago
Actually ethnic Germans also include many people who repatriated from Russia etc. (not that much neighboring). As another commenter said, more migrant population does not lead automatically to better integration.
714.
▲
by
riedel
4y ago
This is first account from meetings I attended as part of the big data value association with W3C presenting on technology convergence for data spaces. Alot was on DIDs which seem to be a pillar of the European Blockchain service infrastruc
715.
▲
by
riedel
4y ago
Only 73% of all people living in Germany have no modern migration background. Beyond that the German nation is relatively young concept (which has proven to be problematic) and there is no homogenic population within the country. Many peopl
716.
▲
by
riedel
4y ago
I am seeing currently w3c people pushing blockchain stuff in European cloud standardization activities. This really worrying me. This explains to me now why W3C seems so broken.
717.
▲
by
riedel
4y ago
I myself work at a university (had the pleasure to be in the working group that drafted our first code of conduct) and I can very much relate to this: that is both sides. IMHO CS professors, however, tend to become drama queens when it come
718.
▲
by
riedel
4y ago
True. However, the situation is not limited to Hesse, but also data protection officers in Rhineland Palladia made the same call: https://www.heise.de/news/Rheinland-Pfalz-Schulen-duerfen-Mi...
719.
▲
by
riedel
4y ago
For anyone else I recommend pdf arranger [1] [1] https://github.com/pdfarranger/pdfarranger
720.
▲
by
riedel
4y ago
I am always quite astonished how bad the default layouters for graphs perform. When I was still doing compiler optimization in the beginning of the 2000s we did not struggle with quite big graphs thanks to cool graph visualizers such as vcg
More ›