1. It will NEVER be finished
Thats the point. Software evolves. Get something that works. If it breaks in prod fix it ad infinitum.

2. MVP. So long as its semi scalable. Basically multi-tenant DB ready and half decent code (not perfect!) then just finish with basic feature set and SHIP IT.

When you get user feedback tweak to thier needs not yours. Let the users decide how it evolves.

Stop changing shit. You are procrastinating and suffering from OCD. Shift your thinking to commercial not technical. No one will apreciate the algorithm you used to compress images (Silicon Valley?). They will however love the fancy menu or drag n drop function and non intrusive notifications and clean UI. Which MVC? No one cares at this point. Keep it lean and you can migrate quickly... later.

If you have a perfect product that you never ship. Youve failed.

If you have a mediocre product that ships, needs fixing post live, gets fixed and improves regularly... and has users. Youve won.

Good luck!

You are all overcomplicating things.

1. Fiber @ 1Gb/s up/down symetrical
2. 5400rpm HDD @ I dunno 20MB (20*8Mb=160Mb/s)

Internet >> HDD

The HDD is the bottleneck. Use iperf or something to confirm RAM vs HDD. Be sure to use Gb LAN or direct to router not WiFi which will unlikely get anywhere near Gb speeds. Also be sure remote server can serve files @ Gb speeds. Etc. SSD might get 500-600Mb/s under very specific conditions... so long as your porn does not judder who cares? :P

Hey.

1. Stop everything and re-evaluate
Why are you stressed? Non payment? Projects too hard? Not enough staff to cope? Bad tech? Shitty clients?

2. Draw out ideal processes, tools, clients, contracts and monetary goals as milestones. Use Trello.

3. Make small changes towards each milestone and no matter what - do not deviate. You will lose clients, staff and maybe money at first.

Once you start re-aligning with a solid structure/plan. The control will reduce stress. Stress nearly always relates to a lack or percieved lack of control.

Use contracts. Over quote on time, NEVER release code before final payment (show clients on local machine or your own server if needed). 50% up front, rest in equal increments at agreed milestones.

Rest. If you are working all night you either have a bad plan, or are not been honest re your abilities.

Dont be afraid to charge alot. Clients would have to pay if they went elsewhere so dont undersell your time.

Run out of chars ;) la la la la la la!!

@GastroGeek

Realise you said boss... so may not be chargeable... still send him my way hehe.

Just build it around PHP Laravel/slim. Make a Document model, add some user privs (public and private and maybe teams), an upload page and a listings page/view page. Use bootstrap with a theme.

Done. 1 day of work each but charge for a week each.... 6K all in. (£5850 as clients baulk? at round figures).

Always say yes... and figure it out afterwards... you can do more than you think...

Then charge per feature afterwards... feel free to send him my way :p

ahhh the old X-WTF gate...

<?php

//story of my life
echo "Shit like every other day...";

If limit of 5 is needed... then:

return n <= 5 ? n*n : -1;

The above does 6 executions/checks regardless of n... pointless if n > 5

Option C

Only another 20+ stages to go :P

But seriously... good luck!

This is why I:

Get 50% up front. Non neg...
Remainder in installments upto go live
Final before deploying to live.

As a bonus. In your ts and cs add a clause to say 30 days post non payment you reserve the right to sell the code to someone else to reclaim losses.

OK... my OCD would rotate the 4th monitor... ya know... for symmetry

response.insert('tumbleWeed');

;)

@NullPointer

Exactly. automate, delegate and reap the managerial title!!! else quit haha.

if its part of the cms... refuse and redirect them.

if you must do it and its not because of a bug you itroduced... charge them!

They will soon learn :)

Solution: work for yourself.

You get to love your job AND love your company... and code naked without getting strange looks which is always a plus.

Its all about the pi zero?

but pi3 > udoo due to support and rapid prototype speed. If the pi3 had 4gb ram and a hardware enabled browser it would be perfect. Shocked that Google have not directly released a GPU enabled/WebGL chrome for the pi... meh. So many project ideas waiting on this.

@dbonev

An uncluttered mind? :p

Agree with majority. Read up on the etymology... datum (day-tum).... brings us to data (day-teh).

When people say (dart-eh). I just shout 'Exception' and stare at them. They soon correct their ways.

My 2c.

Tip. Never relate breakpoints to screen size. its non scalable.

Instead relate it to content. i.e. start small and as room becomes available add a breakpoint and rejig the page... as you increase res... add another etc etc. that way it will work on anything from low res mobiles to outdoor screens.