Hacker News: zzleeper

New comment by zzleeper in "Show HN: Wordif.sh – word-by-word text comparison in color, output to HTML/PDF"

zzleeper — Wed, 25 Mar 2026 05:29:35 +0000

Found it interesting but would have been easier to see an example of how the html looks in the github page!

Show HN: Locro – Fast and accurate local OCR through Chrome's screen_ai

zzleeper — Tue, 24 Mar 2026 04:29:57 +0000

A month ago, @Stagnant posted in this thread about how Chrome ships with a open source OCR tool that is only available from the browser:

https://news.ycombinator.com/item?id=46977802

This looked incredibly useful but sadly there were no Python wrappers, so I followed his instructions and built one.

It's incredibly fast and accurate (I had my doubts but wow!). I tested the Windows and Linux implementations, and I'm sure expanding the wrapper for macOS should be trivial for a decent LLM.

Hope it's also useful to other users, and thanks again Stagnant for mentioning it in the first place!

Comments URL: https://news.ycombinator.com/item?id=47498609

Points: 1

# Comments: 0

New comment by zzleeper in "Some things just take time"

zzleeper — Sat, 21 Mar 2026 17:55:45 +0000

LMK if you finish it, sounds like something my daughter would enjoy!

New comment by zzleeper in "My Random Forest Was Mostly Learning Time-to-Expiry Noise"

zzleeper — Fri, 20 Mar 2026 00:26:31 +0000

That's a refreshing article. Easy to read and I learned a few things!

New comment by zzleeper in "Chrome DevTools MCP"

zzleeper — Sun, 15 Mar 2026 21:46:02 +0000

Another +1, it would be incredibly useful to play with this approach! (and fun)

New comment by zzleeper in "Jeff Bezos wants Washington Post’s newsroom budget halved, productivity doubled"

zzleeper — Sun, 15 Mar 2026 06:22:54 +0000

I'm sure he didn't bought the WaPo to make a profit. More like to have an influence.

New comment by zzleeper in "What happens when US economic data becomes unreliable"

zzleeper — Sat, 14 Mar 2026 17:40:22 +0000

Please don't "both side" this. As much as there is corruption in any administration, R or D, the levels we are seeing now are completely unprecedent and blunt. EG: https://www.cnn.com/2026/03/13/politics/trump-fundraise-emai... (from earlier this morning on the HN frontpage)

New comment by zzleeper in "Elon Musk pushes out more xAI founders as AI coding effort falters"

zzleeper — Fri, 13 Mar 2026 23:07:01 +0000

Wait, what does this imply for Cursor? I DGAF about xAI and will never use their Grok, but I did like Cursor more than the alternatives (even if I'm just running opus 4.6 most of the time).

But now he is poaching the two heads of engineering of a company that's trying to move very quickly, how is that going to affect their speed and success?

New comment by zzleeper in "QGIS 4.0"

zzleeper — Sat, 07 Mar 2026 17:32:33 +0000

I used QGIS 2.x and 3.x a lot when making maps for research papers. But something that always stung was reproducibility. The python tooling was not there compared to what I could do with click-and-mouse, and there was no easy way to transfer my click-and-mouse sessions into an equivalent python script.

Is the situation unchanged? (Maybe a good use for Opus would be to write a wrapper for the python tooling?)

New comment by zzleeper in "US economy unexpectedly sheds 92k jobs in February"

zzleeper — Fri, 06 Mar 2026 18:34:51 +0000

Sad/funny that your comment is at the bottom.

Workers on strike are classified as not employed, so yeah we should ignore that category

New comment by zzleeper in "GPT-5.4"

zzleeper — Thu, 05 Mar 2026 18:58:57 +0000

For me the issue is why there's not a new mini since 5-mini in August.

I have now switched web-related and data-related queries to Gemini, coding to Claude, and will probably try QWEN for less critical data queries. So where does OpenAI fits now?

New comment by zzleeper in "Gemini 3.1 Flash-Lite: Built for intelligence at scale"

zzleeper — Wed, 04 Mar 2026 02:25:15 +0000

Are there good open models out there that beat gemini 2.5 flash on price? I often run data extraction queries ("here is this article, tell me xyz") with structured output (pydantic) and wasn't aware of any feasible (= supports pydantic) cheap enough soln :/

New comment by zzleeper in "Florida public universities to pause hiring new H-1B workers"

zzleeper — Tue, 03 Mar 2026 21:50:31 +0000

If you are in early career (i.e. graduated your PhD within the last 5 years) you are extremely unlikely to get the gifted people visa. The standard approach is to just get the H1B (not the lottery stuff for tech companies but the non-lottery one for hiring faculty at universities). Ask any foreign MIT professor hired early in his career and they went through H1B (and later on, they are more reluctant to move into a place like Florida..)

New comment by zzleeper in "Florida public universities to pause hiring new H-1B workers"

zzleeper — Tue, 03 Mar 2026 21:29:49 +0000

Wow. The amount of quasi-xenophobic comments in this tread is nuts. They are also a bit misguided.

You don't hire a professor at a R1 school just to teach math101. You hire them they can build a research lab or otherwise help to advance the frontier of the field (cancer, stats, etc.). The talent pool in several of these fields is very very small for Americans, because the brightest just go (used to go?) to work to finance or tech. So if you say you can't bring any bright foreigners, you are constraining yourself to a lower talent pool than other countries, and thus will pay a price (in less research, in no foreign students applying and thus no $$ from them, etc etc)

New comment by zzleeper in "Show HN: Scanned 1927-1945 Daily USFS Work Diary"

zzleeper — Tue, 17 Feb 2026 20:53:25 +0000

Thanks for the insights! I'll try Mistral as well.. Gemini worked well for me so far but which model is SOTA is changing quite frequently these days

New comment by zzleeper in "Show HN: Scanned 1927-1945 Daily USFS Work Diary"

zzleeper — Tue, 17 Feb 2026 06:41:43 +0000

That's amazing!

I'm working on a kinda similar project (documenting bank runs from historical newspapers) and also opted for Claude to build a static website. Crazy that the two sites have a very similar look and feel: https://www.finhist.com/bank-runs/index.html . The only big difference is that mine lacks a map, which I should hopefully fix soon (I already have lat and lon and am linking to google maps).

PS: Do you know if mistral works better at OCRing handwritten text than gemini 3? Was planning on going the gemini3 for another project

New comment by zzleeper in "GLM-OCR – A multimodal OCR model for complex document understanding"

zzleeper — Tue, 17 Feb 2026 02:24:02 +0000

That's actually amazing, and might give me a way to use all the cores I have lying around. 2s per page is an insane 600 pages per minute at 20 cores!

Please do open source it, even if you don't do much around it (worst case I can just spend a few million tokens trying to get opus 4.6 to get it to work)

New comment by zzleeper in "GLM-OCR – A multimodal OCR model for complex document understanding"

zzleeper — Wed, 11 Feb 2026 18:03:46 +0000

Surprisingly, I have a few hundred gigs of old newspaper scans so am very curious.

How fast was it per page? Do you recall if it's CPU or GPU based? TY!

New comment by zzleeper in "Amazon CloudFront Global Outage"

zzleeper — Tue, 10 Feb 2026 21:23:59 +0000

I just spent the last 20 minutes trying to debug why my non-www URL wasn't working and my www was. Oh well now I know.

New comment by zzleeper in "The Bash Reference Manual Is in the Epstein Files"

zzleeper — Tue, 03 Feb 2026 16:21:42 +0000

I only see one redaction on page 128 (122 as in top-right of page), and it's just a URL. So there's a rule to redact URLs.