hashhar

21 ms·

181.
by hashhar 6y ago
The governemnt is making large scale detention centers though.
182.
by hashhar 6y ago
True. It's quite telling that as soon as it was morning in India this post fell off the top page.
183.
by hashhar 6y ago
Make sure you have dead comments visible.
184.
by hashhar 6y ago
Sorry, I should have said PPI (pixels per inch) - as calculated by dividing screen resolution by screen area. Thanks for catching that. The rest of the comment still remains.
185.
by hashhar 6y ago
Sadly fractional scaling is imperfect and looks blurry to me (on both MacOS and Windows, GNOME doesn't support fractional scaling (last I checked)). I would love a 5K 27" screen at a reasonable price. :(
186.
by hashhar 6y ago
Nice coincidence running into you here. I did come across your blog while researching! Thanks for that. I agree 100% about the real estate issue. Physically larger monitors with equivalent real-estate don't make much sense for me perso
187.
by hashhar 6y ago
Is this true for just MacOS or do Linux (GNOME specifically) and Windows also work the same way? I don't know anyone with an ultrawide so I can't try this myself.
188.
by hashhar 6y ago
4K at anything below 43 inch is unusable for text without scaling. 27 inch 2K or 43 inch 4K are the sweet spots. Here's how you can check if your monitor needs scaling: - Find out the DPI of your existing monitor and your desired monit
189.
Starburst Data announces $100M Series C led by Andreessen Horowitz(blog.starburstdata.com)
2 points by hashhar 6y ago|0 comments
190.
by hashhar 6y ago
I've been bit by the same issue on ECS. Some stream processing applications restarted and they perform a lot of reads at startup to recover their in-memory state. All other containers on that instance eventually also restarted and got
191.
by hashhar 6y ago
I'm not sure I'll ever need this but this is really useful for the times I get stuck in a situation. Thanks for writing this. I bet we couldn't find this information consolidated in a single place anywhere on the internet. Th
192.
by hashhar 6y ago
No issues in over 5 years. The initial setup was time consuming but now its scripted. The benefits are bleeding edge packages, latest kernel, a lean and mean system (<700 packages for me). Some drawbacks are that getting exotic hardware
193.
by hashhar 6y ago
Thanks for sharing that. I can definitely see myself using the multi-line blocks since they are much better than the kind of things we need to do today. At the same time I'm slightly scared of that magic . It helps that it's we
194.
by hashhar 6y ago
I hate such kind of magic. Very convenient at first but can cause countless hours spent on debugging in some cases. I like how Scala does this via the indent marker and strip.
195.
by hashhar 6y ago
because CRC is actually worse for checking file content collisions (not that MD5 is perfect either).
196.
by hashhar 6y ago
Really great to the point article without any fluff. Lots of actionable insights. The part about treating interviews as technical chats resonated the most with me and has been my personal experience as well - both as an interviewer and inte
197.
by hashhar 6y ago
> The difference is in the culture. As a CEO I would like my client's business to succeed so we can work together as partners, long terms, instead of getting money from CRs and playing "you haven't specified this". I
198.
by hashhar 6y ago
I've recently learnt this trick too but am having a hard time manually diffing stack-traces. Do you know of some tool that can help me analyze samples of stack-traces?
199.
by hashhar 6y ago
I had to write a personal bot way-back-when which listened to all my DMs and mentions and let the people know that I'll get to them before the end of the day and that if I still didn't they could call me up if something urgent. It
200.
by hashhar 6y ago
I see such issues on larger multi-module projects. A very reproducable example would be something like Presto ( https://github.com/prestosql/presto ) or Debezium ( https://github.com/debezium/debezium
201.
by hashhar 6y ago
That is exactly how GitHub is architectured. There have been a lot of time I've been to able to push/pull but GitHub pages was down or their web backend was down. Bitbucket is the worst in this regard.
202.
by hashhar 6y ago
His PhD was in Mathematics AFAIR. Also note that the Wikipedia article is missing a citation on that point and further discussion on the talk page.
203.
by hashhar 6y ago
Just gonna put this here for people who might be wondering. Not too many block-level replication implementations exist. https://www.linbit.com/drbd/ NB: It is Free software, don't be alarmed by the domain name.
204.
by hashhar 6y ago
Iceberg solves more problems than what Hudi does. The biggest win is computed partitions and not having to define a strict partitioning strategy from the start.
205.
by hashhar 6y ago
Yes, we are using it in production (somewhat) - for a a few tables. Our workflow with it is: 1. Stream new data into a partitioned S3 bucket available as a table in metastore. 2. Use Hudi to create a copy-on-write compacted version of the t
206.
by hashhar 6y ago
I can echo this! This is exactly how I feel too - the muscle memory around workspace management and the scratch windows (floating windows that you can toggle in/out of visibility based on a single keystroke) are the real boosters for m
207.
by hashhar 6y ago
Not if you compare on same hardware but the amount of memory itself that you can end up saving you can re-invest it into compute. I've seen massive perf/$ gains for log-parsing and memory-mapped data profiling use cases.
208.
by hashhar 6y ago
Metabrainz and their wiki/documentation/data sites do this. See https://musicbrainz.org/doc/WikiDocs for how they implement it.
209.
by hashhar 6y ago
And the IA followed that rule too but opened up wider access with an "Emergency" library once lockdowns due to COVID-19 were put in place.
210.
by hashhar 6y ago
And how is that different from a physical library.

More ›