Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
oso2k
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
39 ms
·
391.
▲
by
oso2k
9y ago
Ahh. Got you. And easy fix again. Add `$(SRC)` to the pattern rule. css/%.css : $(SRC) scss/%.scss sass $<:$@ css/%.css.map : $(SRC) scss/%.scss @touch $@
392.
▲
by
oso2k
9y ago
Blind includes are a dumb to do (it's one easy way to exploit a source codebase). The better way is to generate your include from, you know, the actual way your codebase is. Use `$(OUT)` to build your include statement. That var say
393.
▲
by
oso2k
9y ago
You don't need to (specifically) teach make about `@import`. It's another scss file and it (hopefully) should be a part of your SRC list. Yes, having something like `gcc -MMd` would be better but that's sass for you. To fix
394.
▲
by
oso2k
9y ago
Recursion does not imply a circular dependency which is most people's biggest concern with Make recursion. A graph with loops is still a graph. A broken graph is actually 2 graphs. If you're careful (and even if you're not)
395.
▲
by
oso2k
9y ago
I don't know if you really understand Make. Even the original article misses point about what Make is really about in recommending `.PHONY` targets. Make is about "making files." Or to be a little more semantically specific
396.
▲
by
oso2k
9y ago
I love the stuff the Archive keeps coming up with. I'm glad I finally started donating to their cause last year.
397.
▲
by
oso2k
9y ago
One other thing I forgot to mention, last week, Blue Apron offered me $30 to rejoin (I already had a $10 credit I threw away). I thought it was interesting and then I saw this HN post. Probably not a coincidence. :/
398.
▲
by
oso2k
9y ago
I agree with you. I used them off and on over 3 months and our family had a similar assessment. We chose the 2 person meal plan because if our (even if nutritionally adventurous) 5 & 7 year olds didn't like the meal, we didn'
399.
▲
by
oso2k
9y ago
Like the MS people, I'm a Red Hatter and I just picked up 15TB last Spring to keep up with my storage needs. I have dozens and dozens of VMs. I consult/work on OpenShift and simulating an HA Cloud environment takes a few hundred
400.
▲
by
oso2k
9y ago
I may be that the moneyball strategy becomes an architectural one; make the walls taller or less conventional like they used to be in the 1920s and 1930s.
401.
▲
by
oso2k
9y ago
Like you mention, that's only been a trend in the last 25 years or so. Prior to that, there was some pride is in being a 30/30, 20/50, 40/20, 50/20 and more rarely, a 40/40 player. All of which could get one
402.
▲
by
oso2k
9y ago
I would say, if it's Millenial Style Snark (gawd I hate to brand all millennials that way but they do tend to sound like the younger halo/gta 5/etc playing crowd) where they say something stupid like "Tests or STFU"
403.
▲
by
oso2k
9y ago
Don't knock snark. It's perfectly valid way to challenge and traces its history to early computer science. It's why you fear goto. It's why I exploit it.
404.
▲
by
oso2k
9y ago
This is pretty cool. It competes well with lz4 even if it is amd64 only. I wonder what an ARM version would look like. Here's to hoping he author keeps working on this.
405.
▲
by
oso2k
9y ago
Full Disclosure: I'm a Red Hat Consultant focused on OpenShift. All of our work is open on github under the OpenShift Origin project, Kubernetes, and other projects. We push most of changes upstream. Doing containers at a scale of 10
406.
▲
by
oso2k
10y ago
This is why I donate. To me this is important. We already lost the library of Alexandria once (or multiple times really until its final destruction). Let's not have our hubris do that again.
407.
▲
by
oso2k
10y ago
You have this posted anywhere? I'd be interested in studying it.
408.
▲
by
oso2k
10y ago
My HP z820 uses almost 300w at load and some fan noise with dual E5-2660s, 128GB RAM, and an nVidia Quadro K2000. I've heard that some people that do audio/video with similar rigs replaced the factory CPU coolers and fans with the
409.
▲
by
oso2k
10y ago
It's interesting. If they wanted to, they could compete with the likes of HP, Dell, Lenovo, and Cisco if they could ramp up production to accommodate customers. I wonder who does their manufacturing on the backend.
410.
▲
by
oso2k
10y ago
HP z820/z620/z420 and Dell T7600/T5600 are great used workstations that support the Xeon E5-26xx V1 series processors at very reasonable prices on eBay. Workstations are $300-$700 (maybe add CPU, RAM, HD, graphics card) shipp
411.
▲
by
oso2k
10y ago
This is great. But please note, some of the graphs presented use a common statistics cheat of logarithmic scale along one or more of the axis. Many times, lz4 level 1 isn't just 25% faster [0] in Compression vs. say zlib level 1 (as i
412.
▲
by
oso2k
10y ago
What makes LLD faster than gold? gold's improvement on gnu ld was focusing on ELF output and removing BFD internal representation. And shouldn't the name be llld?
413.
▲
by
oso2k
10y ago
Mark's a very busy guy at JPL (I'm a former JPLer myself). Making science happen on Mars and whatnot. zlib is super stable and that's largely its biggest benefit as opposed to when it was originally released in the 90s. Bac
414.
▲
by
oso2k
10y ago
I really wish we had TP-style modules [0] information in our object file formats. What seems so simple conceptually is nearly impossible to replicate? Can't figure out why C, C++, Java or C# never got there and why JS is puking all ov
415.
▲
by
oso2k
10y ago
So it is API compatible with V7?
416.
▲
by
oso2k
10y ago
Is this a re-packaged v7 [0] (it's from Cesanta as well)? mjs.c is 477K. It looks like it as v7.c is 475K [1]. It shaves whole entire 1MB from my staticly linked builds (v7 1.9MB vs. mjs 998K), and almost 2MB from dynamically linked
417.
▲
by
oso2k
10y ago
Google's made their attempt a few years ago. It's called GoLD, or, gold [0]. However, its key benefit is in lower link time processing of object files and libraries. Important if you're building apps that take minutes or h
418.
▲
by
oso2k
10y ago
I believe the original Turbo Pascal was written in ASM by Anders Hejlsberg as that was the state of the art at the time [0][1]. Also it had been ported to various CP/M and non-x86 machines prior to becoming an x86 port. Later versions
419.
▲
by
oso2k
10y ago
I'm surprised that no one here seems to have noticed that this is a Tubro Pascal compatible compiler written in JS that writes binaries compatible with 1978 UCSD p-System p-code, and, a p-code VM also written in JS that will run in a w
420.
▲
by
oso2k
10y ago
Not sure what you mean by minimization on failure, but there's a 'classifier' and 'on_fail' method that be might interwoven into the capability you're speaking of.
More ›