Join devRant
Do all the things like
++ or -- rants, post your own rants, comment on others' rants and build your customized dev avatar
Sign Up
Pipeless API
From the creators of devRant, Pipeless lets you power real-time personalized recommendations and activity feeds using a simple API
Learn More
Search - "thrust"
-
This story starts over two years ago... I think I'm doomed to repeat myself till the end of time...
Feb 2014
[I'm thrust into the world of Microsoft Exchange and get to learn PowerShell]
Me: I've been looking at email growth and at this rate you're gonna run out of disk space by August 2014. You really must put in quotas and provide some form of single-instance archiving.
Management: When we upgrade to the next version we'll allocate more disk, just balance the databases so that they don't overload in the meantime.
[I write custom scripts to estimate mailbox size patterns and move mailboxes around to avoid uneven growth]
Nov 2014
Me: We really need to start migration to avoid storage issues. Will the new version have Quotas and have we sorted out our retention issues?
Management: We can't implement quotas, it's too political and the vendor we had is on the nose right now so we can't make a decision about archiving. You can start the migration now though, right?
Me: Of course.
May 2015
Me: At this rate, you're going to run out of space again by January 2016.
Management: That's alright, we should be on track to upgrade to the next version by November so that won't be an issue 'cos we'll just give it more disk then.
[As time passes, I improve the custom script I use to keep everything balanced]
Nov 2015
Me: We will run out of space around Christmas if nothing is done.
Management: How much space do you need?
Me: The question is not how much space... it's when do you want the existing storage to last?
Management: October 2016... we'll have the new build by July and start migration soon after.
Me: In that case, you need this many hundreds of TB
Storage: It's a stretch but yes, we can accommodate that.
[I don't trust their estimate so I tell them it will last till November with the added storage but it will actually last till February... I don't want to have this come up during Xmas again. Meanwhile my script is made even more self-sufficient and I'm proud of the balance I can achieve across databases.]
Oct 2016 (last week)
Me: I note there is no build and the migration is unlikely since it is already October. Please be advised that we will run out of space by February 2017.
Management: How much space do you need?
Me: Like last time, how long do you want it to last?
Management: We should have a build by July 2017... so, August 2017!
Me: OK, in that case we need hundreds more TB.
Storage: This is the last time. There's no more storage after August... you already take more than a PB.
Management: It's OK, the build will be here by July 2017 and we should have the political issues sorted.
Sigh... No doubt I'll be having this conversation again in July next year.
On the up-shot, I've decided to rewrite my script to make it even more efficient because I've learnt a lot since the script's inception over two years ago... it is soooo close to being fully automated and one of these days I will see the database growth graphs produce a single perfect line showing a balance in both size and growth. I live for that Nirvana.6 -
Worked for a team where Scrum was thrust on it to fix/cover up bad management decisions.
Manager is really enthusiastic (to impress his new boss) and gets a bunch of colored post-it sticky notes for stories and shit. Team spends a bunch of time carefully writing stories & etc and arranging said notes on a white board. Come in the next day and 90% of them fell off due to condensation, or such such.
Left the group, with the rest of the senior dev staff, after the "15 minute" standup expanded a minimum of 1 hour each day... and the SM carefully writing down everything everyone said.
A co-dev gave me some career advice - "when the pilots are drunk, it's time to get off the plane."4 -
Holy shit seriously: Fuck MSOffice. Fuck it right in the eyehole.
As desktop software, it's just brutally terrible. On my work mac, it's just sweaty garbage. The latest insult is that on the most recent update, msword stole the default file association from preview.
Libre isn't terrific, but at least it's closed when you close it. For that reason alone, it's orgasmic by comparison.
Because there's justice in the world, my job is not a document-centered one, so I have no real use at all for an office app, let alone the specific macros and formulae that the msoffice versions of these apps provide, so I couldn't give less of a shit about losing functionality.
The headline and main thrust of this rant is "fuck msoffice so hard that it dies of eye-fucking."1 -
*writes programs with variables for arguments*
Runs.. Crashes
*removes variables places exact same code in parameter list*
Runs.. Succeeds.
I don't understand you my GPU -
This is gonna be a long post, and inevitably DR will mutilate my line breaks, so bear with me.
Also I cut out a bunch because the length was overlimit, so I'll post the second half later.
I'm annoyed because it appears the current stablediffusion trend has thrown the baby out with the bath water. I'll explain that in a moment.
As you all know I like to make extraordinary claims with little proof, sometimes
for shits and giggles, and sometimes because I'm just delusional apparently.
One of my legit 'claims to fame' is, on the theoretical level, I predicted
most of the developments in AI over the last 10+ years, down to key insights.
I've never had the math background for it, but I understood the ideas I
was working with at a conceptual level. Part of this flowed from powering
through literal (god I hate that word) hundreds of research papers a year, because I'm an obsessive like that. And I had to power through them, because
a lot of the technical low-level details were beyond my reach, but architecturally
I started to see a lot of patterns, and begin to grasp the general thrust
of where research and development *needed* to go.
In any case, I'm looking at stablediffusion and what occurs to me is that we've almost entirely thrown out GANs. As some or most of you may know, a GAN is
where networks compete, one to generate outputs that look real, another
to discern which is real, and by the process of competition, improve the ability
to generate a convincing fake, and to discern one. Imagine a self-sharpening knife and you get the idea.
Well, when we went to the diffusion method, upscaling noise (essentially a form of controlled pareidolia using autoencoders over seq2seq models) we threw out
GANs.
We also threw out online learning. The models only grow on the backend.
This doesn't help anyone but those corporations that have massive funding
to create and train models. They get to decide how the models 'think', what their
biases are, and what topics or subjects they cover. This is no good long run,
but thats more of an ideological argument. Thats not the real problem.
The problem is they've once again gimped the research, chosen a suboptimal
trap for the direction of development.
What interested me early on in the lottery ticket theory was the implications.
The lottery ticket theory says that, part of the reason *some* RANDOM initializations of a network train/predict better than others, is essentially
down to a small pool of subgraphs that happened, by pure luck, to chance on
initialization that just so happened to be the right 'lottery numbers' as it were, for training quickly.
The first implication of this, is that the bigger a network therefore, the greater the chance of these lucky subgraphs occurring. Whether the density grows
faster than the density of the 'unlucky' or average subgraphs, is another matter.
From this though, they realized what they could do was search out these subgraphs, and prune many of the worst or average performing neighbor graphs, without meaningful loss in model performance. Essentially they could *shrink down* things like chatGPT and BERT.
The second implication was more sublte and overlooked, and still is.
The existence of lucky subnetworks might suggest nothing additional--In which case the implication is that *any* subnet could *technically*, by transfer learning, be 'lucky' and train fast or be particularly good for some unknown task.
INSTEAD however, what has happened is we haven't really seen that. What this means is actually pretty startling. It has two possible implications, either of which will have significant outcomes on the research sooner or later:
1. there is an 'island' of network size, beyond what we've currently achieved,
where networks that are currently state of the3 art at some things, rapidly converge to state-of-the-art *generalists* in nearly *all* task, regardless of input. What this would look like at first, is a gradual drop off in gains of the current approach, characterized as a potential new "ai winter", or a "limit to the current approach", which wouldn't actually be the limit, but a saddle point in its utility across domains and its intelligence (for some measure and definition of 'intelligence').4 -
Quote of the day:
Some are born developers, some become developers and other have development thrust upon themselves!1 -
Family - "You might be a great coder, but if you dont have a degree, you wont get respect"
Overall, my brother and family are very supportive. only thing i dont like is their strong thrust to get me graduated and pass the exam, despite my negative interest in the course i am enrolled into.
I am into programming, not because of money, that is flowing around it, or the present day trend. Infact programming was not my field of interest, even though I was always interested in computers and machines. My first choice was to become a mechanical engineer. But i got enrolled into Electronic communication course.
Only after sophmore year, i started getting interest in programming.
Programming, to me is a very powerful tool. You can create what you want.
Programs, when works, Feels like magic -
Some pigs just don't fly, even with sufficient thrust.
Been struggling for days trying to get my IntelliJ-on-Linux-VM setup running on Windows, and it's abysmally slow. Latencies for the UI are counted in seconds.
Screw it, I'm reinstalling my entire dev stack direct on Windows. I love Linux, but I will not use it when the price is that I cannot get anything done.8