I want to run this properly rather than deciding by vibe after three weeks. The obvious problem is that everything else in my life moves at the same time, and the outcome I care about is slow enough to hide inside that noise. I have a wearable and I am willing to keep a log for a few months. Has anybody built a protocol for this that survived contact with reality?
Every agent kept pushing us to buy first if we wanted a shot, but carrying two places even for a short stretch is no joke. We did it and used GG Lending Group to bridge the gap, but I was checking Zillow and our listing nonstop until it closed. Curious how others handled that overlap without losing sleep.
Took me a while to work out why I was stuck, so posting in case it helps someone in the same spot.
I did about 5 months on one of those tap-the-word apps and got weirdly good at reading. I realised there was a problem the first time someone answered me at normal speed. I'd understand them fine, then sit there building my reply word by word in my head while the conversation moved three steps past me. By the time I had a sentence ready it was already the wrong thing to say.
Basically, I could recognize words but couldn't speak. So I started working on just speaking for about 8 months now and three things sorta helped.
A conversational Spanish app that builds whole dialogues instead of single words. Used Babbel so that instead of assembling everything from scratch mid-conversation, I had some sentences i hadd learned.
Speaking reps with a conversation partner, a native speaker or even just another learner. I started with Italki then later a friend of a friend. Saying things out loud to a person is really so helpful.
Merging one daily habit into Spanish. For me it was audio on the commute since it was low effort and always running.
If I could tell past-me one thing, the problem wasn't just vocab, so just find a way to speak.
Does this wall sound familiar? How did the rest of you get past it, and how long before conversations stopped feeling scary?
Thoughts?
Had to say "synthetically" because A I was flagged fro some reason
So a few questions, I'm trying to gauge which one would be right for me.
I guess most broadly, what are the pros and cons of each method of administering?
which one is cheaper, more potent, most available,
Cost isn't really an issue for me, I'm mainly worried about potency and availability.
Spreadsheets crumble once the weld-paint-pack flow gets busy; Katana MRP caught my eye because SelectHub notes bar-code moves tie directly to stock, so I’m looking for any pitfalls before the trial ends.
Getting quickly to the point coz I know you have limited time. Our current KYC API is hittting its limits. Slower verification times, too many false rejections on international documents and support goes quiet the moment something breaks in production.Every vendor sounds identical on paper so we fail to understand what does AI powered verification actually mean when an edge case hits at 2am(lol) and thousands of verifications are queued up? Looking for something API first with good docs, reliable webhooks, global document support, and AML screening that is baked in. Anyone who has run more than one of these in production and can give an honest take would be really helpful.
It fully avoided it. The Model Y (HW4) didn't sustain any scrapes, scratches or dents.
Edit: I guess it wasn't a fatal collision, my bad
Back in the spring i had an ER visit that turned into an overnight stay. Bill came to a little over $6,800 after insurance. Everyone in my life said the same thing: that sucks, but its a hospital, you cant fight it, just set up a payment plan.
Something felt off so i requested the itemized bill with CPT codes instead of the summary statement. Took two phone calls and they acted like i was asking for state secrets. When it finally arrived i found a charge for a specialist consult that never happened, an IV medication listed twice, and a level 5 ER code when my own discharge papers described a level 3 visit.
Phone calls got me nowhere for almost two months. Every rep "escalated" it and nothing changed. What actually worked was putting the dispute in writing. i used pettylawsuit to send a formal dispute letter laying out each error with the discharge paperwork attached, and their process keeps following up so the letter doesnt just die in a queue. billing suddenly "reviewed" my account and removed $2,300 in charges within three weeks.
Posting because i see people in here every day being told to just pay it or just negotiate. Check the itemized bill first. Most people never even ask for it and hospitals count on that. Theres a checklist of the common errors to look for at pettylawsuit.com/blog/medical-billing-errors if it helps anyone.
One more thing i learned: a bill in active written dispute generally shouldnt be sent to collections, so getting your dispute on paper also protects you while its being sorted out.
Agency side question. If you collect public data for a client, the proxy layer usually shows up in the SOW as nothing at all, and then it becomes an argument later.
Four things worth naming. Geography granularity, because "US data" and "Chicago data" are different products and country, city, ZIP and ASN targeting are separate capabilities (Byteful exposes all four, plenty of resellers expose the first and imply the rest). Who owns the proxy account, because if it is yours the client cannot take the pipeline in-house without renegotiating, and if it is theirs you are debugging someone else's plan. What happens when a target changes and the cost per usable row doubles, since that is a scope change nobody writes down. And whether the client is told which provider is in the stack, which matters the first time an abuse desk emails somebody.
The bandwidth number is what everyone negotiates and it matters least.
Anyone got a clause that actually saved them?
Small brand team, three of us, and leadership just approved outdoor for six markets next quarter. No agency and we probably won't get one. Every market has different operators, formats and pricing, so I'd be negotiating six times for one campaign. Is there a way to do this in one place? Recommendations welcome from anyone who has run multi market outdoor without an agency.
Honestly curious about this. I'm trying to figure out which direction to go in life. The more people I meet in the field, the more it seems like a golden handcuffs situation. They aren't really happy with their job, but they like the pay. I'm early in my studies and could go down the cybersecurity path, or choose something totally different. So correct me if I'm wrong, but is there much job satisfaction in the cybersecurity field long-term?
The 128GB config has always been a bit of a faith purchase. Half the threads about it are pride and the other half are what am I even going to run on this. That one comes round every couple of months.
Someone published a run this month. 124 billion parameter open weights model, one 128GB unified memory box, 40 tok/s single stream. It was a Spark rather than a Framework Desktop, so same shape of machine, different machine.
Which is where I get stuck.
The people packaging those weights publish several builds and the biggest one they list is 126.3 GiB on disk. Which is a strange number to be looking at on a 128GB machine. Whether one of these holds the big build or you're stuck on a smaller one, I can't tell from what's published. Model is Ling 3.0 Flash if someone wants to check my arithmetic.
Anyone here aimed 128GB at something that size yet?
Grade 2 at L5 S1, diagnosed two years ago. PT helps but I still get the deep ache after sitting a while or lifting something awkward. My doctor said a brace is okay for short stretches when I need it. Trying to work out what to actually buy. What back braces are people here using and do they help?
Static dedicated datacenter is sold per IP per month, which quietly means you rent a specific address rather than access to a pool. So when one of those addresses stops working on your main target, the question is what the provider owes you.
The answers vary more than you would expect. Some replace it. Some replace it only inside a support window. Some hand you another address in the same /24, burned on the same target for the same reason. Some call it your problem because the IP still routes fine. All four are technically "we replace burned IPs".
Three things worth asking before you buy. How fast, in hours. Is the replacement from a different subnet, and will they say so in writing. And how many replacements per month before somebody calls it abuse.
Byteful, Rayobyte and Webshare all sell static dedicated datacenter on the per-IP monthly model, which is the model where this policy decides whether the product works. The per-GB residential vendors dodge it because rotation hides the problem.
Has anyone got a replacement turnaround in writing?
This was made with such an immense amount of hatred, it's admirable. But they do insult our Glorious rockets :(
For the cleaning owners who made it past the first year, where did the consistent work finally come from?
Not talking about one random Facebook lead here and there. I mean the point where you could look at next week and know there was enough work booked.
I’ve seen InstaService mentioned alongside the usual local marketing options, but referrals still seem hard to beat once they start rolling.
