244 karma · joined March 2, 2013
Other great examples (though in a whole other league) are the phoenix project and the unicorn project, where the characters learn along the way.
Of course the learning and story is "guided" by the author. Every story ever written was.
Got any suggestions on the topic?
Br
You should definitely have proper time lines, that has nothing to do with full stack developers.
Downtime? If downtime is an issue for your system, most likely you can build it so it can be live with some parts off and do maintainenance during working hours. (Redundancy, etc)
Integration and dB teams, imo, is an antipattern.
But what I really did not agree with with your original comment was that managers look for full stack devs because of cost.
I look for the right people, and I build my teams with a mix of frontend, backend and fullstack devs. It has nothing to do with cost, it has to do with building efficient and dependentless teams.
Imho; a fullstack developer is a person who can develop all pieces of an application (or game, or whatever) and put it in production, maintain it, improve it.
I.e. design the database or other persistence, implement the layers that manages the data, the business logic and the user interface.
It can be very different from different applications, of course, since some may not have an ui, or a persistence requirement.
It is a person who not only knows how to write code, but also understands how code communicates with services and humans.
A fullstack developer understands OS and computing in general.
When I build teams, I like to have a mix of fullstack, frontend and backend developers because the experts (i.e specialized Devs) often handle the nitty gritty details of front- or backend, whereas the fullstack glues together the front-, back- and ops.
Usually fullstackers are the kind of people who say "oh, interesting problem, let's explore" rather than "I don't know this technology, I am a (Fe|be) developer.
You will rarely hear a full stack developer say "I'm unable to continue, because I have to wait for the (dba|integration team|ops team)". They will solve their problems and often has enough experience to do it well.
Preferably cache as close to the consumer as possible, for example using a cdn.
If you can't afford/use a cdn, then enable equivalent caching in your load balancer. (Nginx, haproxy, probably Apache, can all cache very well)
I have only experience with nginx, and there you can configure the caches to serve stale data if upstream is overloaded, and you can set it to refresh cache in background. It helps alot!
If you deliver a resource from nginx memory cache, there is not much more than transfer times in play!
In your services, precalculate as much as possible.
For example, don't gzip/brotly/compress on the fly, do it at build time when possible.
Same with dynamic data from a backend that needs to be preprocessed. If possible, do the preprocessing "outside" of the request handler, in advance.
When it comes to actual performnce of services, in my experience go and java delivers a magnitude better raw power then nodejs. (It might be me.)
Unfortunately, if you need to do eg serverside rendering of js templating system (vue, react, etc) youre pretty much doomed to use nodejs for serverside.
In this case, make sure the serverside generates data that is cacheable and "same for large chunks of users". This means avoiding cookies on requests that don't need them, since cookies.in general renders the payload uncatchable.
Caching is also a reason to avoid graphql. Imo, when you make graphql queries cacheable, you lost all benefits of graphql.
When it comes to data used in backend services, there as well you can use caching layers in smart ways. For example, delivering product data might require up-to-date stock info, but lots of other info (name, description) is changed rarely.
In this case, cache the stable data longer and the dynamic data more often.
When it comes to horizontal scaling, remember that in-memory caching done in backend services is local to that service instance. So if scale up, you might get more requests to the dB.
When it comes to dB, make good choices and limit heavy joins, transactions, etc in real/request time.
By good choices, I mean: it might be better to store session info in redis than SQL. I'm currently using an ecom platform where they store everything in SQL dB. Sessions, shopping carts, etc. Needless to say, the dB server needs to be beefy. They could've easily put sessions and shopping carts in a scalable kv-storage without drawbacks.
If you can avoid sessions (or server state), do it. If you can't, limit their use and size.
Once a clever developer at a previous gig decided to put "recently visited products" in session of user. It was not optimal. Lots of data going back n forth. Filling memory. Etc.
When processing requests, it might be a good idea to have knowledge of when to use lists, when to use hashmaps, etc, so you don't end up in crazy waste of resources.
If a request takes 10 or 20 ms of CPU makes a big difference if your users/rps increase.
Also, try to make your services horizontally scalable from beginning. Imo, much easier to do right from start than to fix afterwards.
Last, but not least, load test you apps using a tool like jmeter, locust or (possibly preferably) goose.
Have fun :-)
(Edit: name of load test tool)
Setup your servers using ansible, puppet or similar to lower the requirements for redundancy and backups (if you cant take a little downtime in case of accident)
If you're Europe based, one suggestion is to get two ~$50 servers from Hetzner and run all services in plain docker. (i.e. two servers for redundancy)
Depending on your setup and requirements, it might even be enough to have one server + a cheap CDN and in that case you might even be able to run the server at home or a friends office.
But it's equally nice to use Firefox, turn off tabs, and let the window manager do it's magic (uh.. with a little help form alttab if youre on osx;)
It's great with managers that have domain knowledge, but having domain knowledge does not make you a good manager.
(Unlike the equivalent in KDE that lets you filter by window title by just typing)
The computer to me is a tool and thus switching window / focus should be swift and efficient.
KDE lets me do that out of the box, other DEs needs addons it seems..
The problem with original macos cmd-tab / and in-app window switching is that it gives you two levels of switching. For the same reason I'm not a fan of tabs in browsers.
I don't want to keep track of which app a window/page is opened in. I want to get to it as fast as my thoughts realize I need to get there.
Most often I work in 2-3 windows in different apps, so quick jumping back and forth is crucial and alttab is a simple and fast way of doing it.
It would have been interesting with restore times as well.
Quite important if it will take you an hour or a day to restore..
L
When you get your pay, put aside money first and save for dedicated budgets.
Decide what budget you think is required in each slot.
Example: I want a house saving for min $30000 for unexpected and expected maintenance.
I want a minimum of $8000 for vacations.
Etc
Then, save as much as you can into those accounts until you reach their goal and then keep saving into them, but maybe lower the amount.
In the beginning (and if your out in time, this will be while you're DINKing) it might be rough, but it will be easier as your buffers stabilize.
When fixing the house, don't steal money from your travel buffer. And absolutely never take money from house buffer for travels. :-)
I think a lot of people spend way to much on luxury, like eating out or having three movie streaming services instead of none or one. And then they realize that "oops, no money left for savings this month neither"
Survival, savings, higher standard, then luxury, not the other way around.
I by far prefer KDE.
I think things work pretty much as well on Linux/KDE as they do on Mac/win. The apple mantra "it just works" is just not true.
With Linux/KDE I get extensibility and can make the platform streamlined to my use cases. That said, it's not perfect, but the amount of wtf/day is less on Linux for me.
One simple example is the file dialog that is a disaster in gnome and Macos. Windows so so, KDE it's good.
The goal to simplify things for end users that runs deep on Macos and gnome I think makes things more unusable than simple and clear.
The quick navigation between windows is another example where I really don't understand how people can work efficiently on Mac. On Linux I can easily jump between windows, have virtual desktops with organised applications. On vanilla Mac you have cumbersome keyboard navigation and have to switch to one app before you can switch to the right window of that app.
I'll stop ranting now.
I'd love read the tldr
Using rest, it's dead easy to enable caching on load balancer or CDN level with background updates.
But with graphql- it's not so trivial. Youll have to do workarounds by caching POST requests (which imo is bad practice) or use GET requests, which might not work because the URL gets too long.
Or use named queries, but then - what did you gain compared to rest?
Another problem is that you expose internals and may more easily make yourself vulnerable to overload attacks.
My two ören
KDE is by far my preference and in general I don't think neither windows nor macos has fewer problems despite the price tag.
After hearing the "it just works" mantra of apple users for many years I was surprised to find I had at least as many glitches on the Mac as I did on KDE (win 7 was better, 10 has more problems ime).
(I use my computers for development and sysadmin, not gaming or art)
This is not optimal for documents to be read on screen/mobile devices since they handle zooming, changing text sizes etc poorly.
In a way I can agree with you that application delivery would allow document distribution as well, but I think the benefits of a unified standard for this that focuses on content rather than presentation is a key feature of the web that we should not loose.
I.e. publishing documents for reading and making applications for interactions.
I think the browser is a pretty good tool for browsing information (think wikis, blogs, news, etc) but we are shoehorning applications (think project management tools, calendars, mail..) into them.
I would like to have them separated so one tool can be optimised for documents and one for making applications.
Connect to an existing session and just "continue where you were" is not easy though there might be unmaintained solutions.
If you use ssh forwarding and for some reason gets disconnected, the application dies.
I'd love for you to tell me I'm wrong :-)
The way rdp works on windows is the only thing where I think windows is superior.
Vnc, TeamViewer and similar are not remotely close and they will leave your remote screen open for anyone to see.
I assume you've noticed the injected js and other fun things people are sharing?
Is there an easy way other than copy paste to rectify this?
Check your VPN config, it should update systemd settings automatically when connecting/disconnecting.