Posted on

On naming (in software)

This is a piece of a conversation that took place elsewhere, a commentary that I think people might find generally useful – therefore, I’m posting it here.

Just for the record, the reason `withassets` is one word (I understand it’s not very English) is because (1) “with” is one syllable, and (2) it’s a single token.

On counting syllables: I prefer names that are short, while still being descriptive. Previously, I named something “more notes” when some reviewers favored “additional notes”. Generally, I think 2-3 syllables is a good length for a name. Good examples: facebook, instagram, google, sprokets. Bad examples: wikipedia, asset pipeline.

On lack of underscores: an underscore is safer than a dash (which is a mathematical operation), but it’s still a special character that separates tokens. I saw a client database where table names were just a-z letters, no capitalization, no other chars. While other readers ridiculed that choice, I found it interesting. `withassets` is a sub-environment so to say, and it’s one thing. It’s not separable into “with” and “assets”. If there is ever a need to have `development_withcache`, `development_withssl`, splitting environment name by underscore clearly gives you the major and minor parts. Introducing an underscore in the middle of a token may require more complex parsing.

Posted on

Deepfakes?

They were made illegal fairly quickly. A few Reddit subreddits were shut down over digital privacy violations as soon as they were created. Anyway, how does one get onto the deep fake train?

Turns out this is the software suite that’s used “95% for the time” for deep fake creation: https://github.com/iperov/DeepFaceLab

And here is a tutorial: https://mrdeepfakes.com/forums/thread-guide-deepfacelab-2-0-explained-and-tutorials-recommended

Of course, this is a reminder to myself as much as anything else, to go back to this and read up and do a trial run, when I have the time.

Posted on

Definition of Priority Codes ( P0, P1, P2, P3, P4 )

I thought this topic is clear, but I’ve run into this again very recently. Since so much conflicting information is going around, let’s iterate once again on what the priority codes should mean. I’m drawing this originally from Jira definitions  – I think that’s where i saw it first, and it is the most sensible.

  • P0 – the work stops, site is down
  • P1 – unblock someone else, required to be done before other things
  • P2 – ordinary flow of work
  • P3 – nice to have, but not required
  • P4 – informational only

I recently saw another definition coming from a more “senior” individual, and it almost made sense, until it didn’t. At which time I stopped and re-evaluated the definitions for myself, once again. The faulty definition was:

  • (DO NOT USE)
  • P0 – the work stops
  • P1 – the task affects customers
  • P2 – the task affects customers, but there is a non-technical workaround solution
  • P3 – the task doesn’t affect customers
  • P4 – informational / never used
  • (DO NOT USE)

The problem with the faulty definition is that everything is P1, and nothing can really be escalated to P0. P4 is completely useless, and junior PM’s and developers would usually create P3’s (the default in Jira), positioning themselves for a need to clarify or shift priorities, etc.

Another thing to note. The priority definitions are technical. They aren’t written with customers in mind – there are multiple layers between devs and customers, and the priority codes are for developers more than anyone else.

The default code is P2. If the work doesn’t need to be done, save the resources and time and don’t do it; change the priority to P3. If there is a change that a task needs to be done sooner rather than later, bump it to P1. And the P4 is effectively never used, because informational stuff resides in wikis and other communication channels..

Posted on

Good-Bye, Google

How I am planning on controlling the products and services I use more strictly.

Relatively I’ve gotten a new phone, a moto x4, and since I’m on Project Fi (by Google), this phone ended up being a Google phone, too. I only really got it for one reason: it has a dual rear camera, allowing me to take nicer pics. I can use that, for my Instagram. However! After 3 or so months of using it, I have to say I’m ready to give it up, and finally switch to an iPhone.

Now, why? The short answer is that I really really don’t like the Google assistant. It tells me when I get home, it tells me when my credit card is due, it knows everything about me. My photos are backed up (poorly) on Google cloud – so I can neither get them back easily, nor have any reasonable privacy. And I use google way too much. It has all my data and when I type in a search query in Chrome, it auto-completes it for me, so very often instead of searching for what I want I search for something else – the nearest-popular autocompleted sentence. I don’t feel comfortable with that at all.

Now, Google will still have my data, and true privacy is impossible to achieve. But I should make a conscious choice to at least attempt to mitigate the risk and the problem of privacy. Just look at how much trouble Google+ has been, or how much trouble Facebook has been! Oh, I quit Facebook quite a while ago, I have an account but I don’t rely on it for communication or any part of my social life. However, real tools like the email, calendar, and the physical sellphone are harder to own as private.

Another reason for me to dislike google is that they charged me $1800 according to some 5-year-old contract, because someone used some API keys that were under my name. I’m still recovering these monies, actually. Obviously nobody reads the contracts they sign, but also – let’s sign fewer contracts, and let’s actually avoid unnecessarily giving away control.

In general, I don’t feel comfortable how pervasive Google’s services are. I want to use it less, not more. So I’m in the slow, gradual process of abandoning google services. I feel that email may be the hardest to abandon.

Posted on

Live Systems

Everything changes. And change works – life itself has proven that.

In technology, you would think that since computers are very good at repeating something very exactly over and over again, things in the digital world sometimes wouldn’t change… But they do. They have to. There is somethign called digital rot – everything rots if it doesn’t change. Relationships rot (try not talking to a close friend for a week), tools rot (who uses myspace anymore?), and if a digital offering doesn’t undergo continuous change, it gets left behind and dies off. So you have to change, and your products and services have to evolve, or be left to “die.”

What twitch.tv teaches us also is that a good media service has to be updated very regularly. Preferably every day. If something is not happening in a channel for a day, it’ll start losing popularity. So you have to continue innovating, continue generating content, and be more varied and precisely match what current trends are set by the market. It’s a hard job, but it’s the only way to survive (and prosper).

The good news is that if you are very new, if you’re just starting out, the continuous change aspect of things doesn’t disadvantage you, on the contrary! As something new on the field, your product/service has room to grow, you haven’t figured everything out so there is very much room for change, and change is often good. So it’s a natural way to level the playing field: the newcomers have this edge over the old-timers, in that the newcomers will necessarily change.