This might actually work!
Crossref is hiring a Director of Technology
A templating system using the file system for inheritance
Way back in the early days of the web, around 2004, I wrote a templating system that used the file system for inheritance. I think Fred Toth originally conceived of the technique.
In the directory /A/B/C you place the template M with content
Hello [%include N%]
You then have the templating system expand /A/B/C/M. It would execute the directive [%include N%] to include the template N by looking up the directory tree, in order, /A/B/C/N, /A/B/N, and /A/N, and using the first N it found. You would place common templates (eg headers) and default content (eg company name) in the upper directories and "override" them in the lower directories. It worked really well for the mostly static sites my department was creating.
I have not seen something like this elsewhere. You can, however, achieve the same effect by manipulating your templating system's template search path per output document.
The system came to be called Trampoline and it has a Perl and a partial Java implementation. The implementations are in the Clownbike project at Source Forge. None of the templates Clownbike used made it to Source Forge, unfortunately. Those became the proprietary web sites our customers were paying for. Galley, an internal project, seems to have some.
I have no idea if any of this code still works. I am sure to be embarrassed by the code's quality! Some quiet, rainy day this winter perhaps I will try running it.
Red Indian Pipes
On a walk this weekend I saw a red Indian Peace Pipe. Neither I or my wife had ever seen one before. Apparently, they are not common, but also not rare.
Setting a Mac's default email client
I mostly love using Macs, but sometimes the conviences provided are not. I needed to change my default mail client to Microsoft Outlook. You set the default mail client within Apple's Mail app's Settings. However, you can't access Settings unless you first configure an email account! Since I don't want Mail to touch a actual real email account I ran these mail services locally using Docker:
docker run virtuasa/docker-mail-develThis enabled me to configure Mail to use "debug@example.com" and the local POP server. And now I can access Mail's Settings to set the default mail client to Microsoft Outlook. I really do feel for all those users without there own System Admin.
MSCHF's “Not Wheels” and Osprey's Gaslands
Examples of hardiness
I walked at Tippacansett this past weekend and I was drawn to these examples of hardiness. Perhaps I was thinking, consciously or not, about still being an individual-contributor software developer at 60.
Using data from external files
I am working with some code that processes CSV files. Each row corresponds to an existing record and the record is updated in response to the column values. This is not an uncommon task. The existing code implements this in an also not uncommon way by intermixing row parsing and record updating. For example, assume we are updating Foos
class Foo attr id, :location, :valuation # ... endA typical intermixing is
row = [...]
raise "unusable foo id" if row[0].blank?
foo = Foo.find(row[0].to_i)
raise "foo not found with id #{row[0]}" unless foo
raise "unusable town location" if row[1].blank?
location = Location.find_by(town: row[1])
raise "location not found with town #{row[1]}" unless location;
foo.location = location
raise "unusable valuation" unless values[2].to_i < 10_000
foo.valulation = values[2].to_i
While this initially seems like a reasonable approach it quickly breaks down as the number of columns increase, column format is non-trival, and there are column (or row) interdependencies. But the more significant problem is that the parsing and the updating can't be tested individually. This makes the test harder to write, understand, and maintain.
It is always better to first parse the raw data, validate it, and then use it. Eg
class Record
attr :id, :town, :valuation
attr :foo, :location
def initialize(values)
raise "unusable foo id" unless /^\s*(\d+)\s*$/ =~ values[0]
id = $1.to_i
raise "unusable town location" unless /^\s*(.+)\s*$/ =~ values[1]
location = $1
raise "unusable valuation" unless /^\s*(\d+)\s*$/ =~ values[2]
valuation = $1.to_i
end
def validate
foo = Foo.find(id)
raise "foo not found with id #{id}" unless foo
location = Location.find_by(town:)
raise "location not found with town #{town}" unless location
raise "valuation does not match the minimum" unless valuation >= 10_000;
end
end
# read the raw data
rows = [[...], ...]
# parse and validate the data
records = rows.map do |row|
record = Record.new(row)
record.validate
end
# use the data
records.each do |record|
record.foo.location = record.location
record.foo.valuation = record.valuation
record.foo.save
end
This parse, validate, and use approach is approporate for all cases where you are bringing data from the outside into your application, no matter what the outside source.
ps. These small, helper classes are your friends. Prefer them over your language's hash primitive as they provide great control. Most languages have efficient syntax for creating and using them.
Short list of useful Mac utilities
I don't add a lot of customization to my Mac, but these I have found to be very useful:
Doll to show messaging apps' icon and badge in the menu bar. (I turn off all notification sounds and banners.)
cd to app Finder toolbar extension that opens the current directory in the Terminal.
The recent discussion Taking command of the Context Menu in macOS on Hacker News has good advice on configuring the Finder's context menu with the ContextMenu, Automator, and other tools.
The Paris Olympics' pictograms
Remain in control of your runtime
Several years ago I wrote Remain in control of your search. When you directly provide to your users a mechanism that your application depends on for its runtime then you immediately lose control of your ability to evolve the application. The recent speaker on the Stack Overflow podcast spoke about how ScriptRunner was implemented in Groovy, and it directly provided Groovy as the user's scripting language. When the application's evolution required updating Groovy the developers incurred the burden of ensuring that all the user's Groovy scripts would continue to operate correctly as is. Scripting languages designed for embedding usually provide a sandbox mechanism for limiting access to runtime resources, but few (none?) limit the language itself. It would have been better if ScriptRunner's design used a Groovy-like language that they translated to the runtime language. Having done that they would have been free to update Groovy, incurring only the burden of ensuring the translation continued to work.
Maintenance, evolution, and technical debt are different
I was listening to a recent Stack Overflow podcast on technical debit and was getting progressively annoyed. The cause of the annoyance was that the speaker was lumping all changes that were not feature changes as technical debit. The changes he was speaking about are better characterized as either maintenance, evolution, or technical debt.
Maintenance is bug fixes. The original implementation is deficient and needs to be repaired. This work is only related to the original design and current feature set.
Evolution is change needed due to alterations in the runtime environment. This work is only related to the original design and intended runtime environment. For example, you need to update a library and that results in a need to update the original implementation.
Technical debt comes about when the original design is unable to accommodate a new need. For whatever reason, it is no longer suitable.
Too often "technical debt" is used to mask an organization's lack of timely maintenance and planned evolution. Don't do that.
Naming things once
How is x = %i[a b c] better than x = [:a, :b, :c]? I find these kinds of shortcuts available in Ruby, and exploited to the maximum in Rails (et al), a severe detriment to understanding and exploring the code. Thankfully, JetBrain's RubyMine is almost always successful handling these pointless alternative codings. Ruby & Rails: Preventing successful grep-ing since forever.
Update: This is a somewhat related discussion on Hacker News, Greppability is an underrated code metric.
Understanding Patterns of Project Behavior
Why can't we validate a branch locally?
I have been on several projects where a developer's version control branch needs to pass continious integration (CI in CI/CD) on a remote service before it is eligible to be merged. The implication here is that a successful remote test run is sufficient validation for acceptance. But given that the remote CI environment is also unlike the production environment I have been wondering why do we accept this validation over a local validation?
It seems to me that an organization should strive to enable local validation over remote validation if only for its cost benefits. Central CI environments are costly to operate. A common consequence of this is that these costs are reduced by having several test runs share an environment. Tests fail not due to a developer’s negligence but to the missing isolation. Validation now requires multiple runs in the hope that the next one will avoid all the conflicts and pass. It is CI wack-a-mole style, a style that costs too much and delays feature releases.
All developer machines are capable enough to run CI in the background while development happens in the foreground. The only thing that is missing is a signifier that authenticates the successful run. Eg, a value tied to the code commit, the successful test run log commit, and the environment commit.
SQL Antipatterns: Avoiding the Pitfalls of Database Programming
Sol Rashidi on delivering an AI product
This interview Your AI Survival Guide • Sol Rashidi & Joe Reis • GOTO 2024 is a useful discussion on delivering an AI product.
I read Sol Rashidi's book Your AI Survival Guide. While it is focused on delivering AI projects, it is also her distinctive take on project teams and delivery in general. Chapter 4, "How to Start", is very good and worth reading alone. I especially liked the quantitative readiness assessment as a means of selecting whether the business should be attempting an efficiency, productivity, effectiveness, expert, or growth AI strategy.
Subscribe to:
Posts (Atom)