Friday, March 13, 2026
Home Blog Page 128

PHP language nonetheless related, advocate insists

0

A longtime main programming language for net growth courting again to 1995, PHP has misplaced the highlight to languages resembling Python and JavaScript lately. Nonetheless, a PHP specialist at PHP software program vendor Perforce Zend is stressing the continued significance of the language.

“Is PHP nonetheless related in 2026? Quick reply: Sure, and it exhibits no indicators of going anyplace,” stated Matthew Weier O’Phinney, principal product supervisor at Perforce Zend and OpenLogic, in a January 15 weblog put up. O’Phinney developed net functions on the PHP-based Zend Framework even earlier than its public launch and led the Zend open-source mission from 2009 to 2019. PHP, he stated, has been the silent workhorse of the fashionable net for greater than three many years. “The truth is, many customers are interacting with PHP daily with out realizing it,” he stated, citing PHP utilization within the Drupal and WordPress content material administration methods. Frameworks resembling Laravel and Symfony additionally use Zend, he added. “From private blogs to complicated enterprise methods, PHP’s utilization stays widespread, whilst newer applied sciences emerge and develop,” O’Phinney stated.

“Whereas it’s true that PHP utilization has declined barely lately, it stays the preferred alternative for server-side languages by a large margin,” O’Phinney confused. And with developments in PHP 8.x, efficiency isn’t a bottleneck for PHP net functions, he stated. The JIT (simply in time) compiler and enhancements to the Zend Engine be certain that PHP handles high-concurrency requests effectively, he added. This previous November noticed the discharge of PHP 8.5, that includes an extension for securely parsing URIs and URLs. PHP additionally has confirmed itself extremely adaptable to cloud-native and containerized deployment, O’Phinney added. “The language simply integrates with containerization instruments like Docker, empowering groups to construct light-weight, remoted PHP environments which might be constant throughout growth, testing, and manufacturing phases.”

Google Antigravity: AI-First Growth with This New IDE


Google Antigravity: AI-First Growth with This New IDE
Picture by Editor

 

Introduction

 
For the final couple of years, the unreal intelligence (AI) revolution in coding felt like having a really quick junior developer sitting subsequent to you. Instruments like GitHub Copilot or Cursor had been wonderful at ending your sentences, however you had been nonetheless the one holding the steering wheel for each single flip. You needed to copy-paste snippets, repair the imports, and manually run the assessments to see if the AI truly knew what it was speaking about.

We’re formally transferring previous that. Google Antigravity marks the start of the “agent-first” period. It is not only a Copilot; it’s a platform the place you cease being the typist and begin being the architect. Google selected the identify “Antigravity” for a really particular purpose. In conventional growth, there’s a large quantity of “gravity” — the heavy, tedious weight of establishing environments, debugging boilerplate, and leaping between the terminal, browser, and editor. Antigravity is constructed to supply liftoff. It’s designed to drag you out of the weeds of low-level syntax and elevate you to the next, task-oriented degree.

As a substitute of writing code line-by-line, you present the imaginative and prescient. The built-in growth surroundings (IDE) autonomous brokers deal with the development, the testing, and the verification.

 

Google Antigravity Landing PageGoogle Antigravity Landing Page
The Google Antigravity touchdown web page | Picture by Creator

 

The story of Antigravity is definitely a narrative of Google going again to its roots. On November 18, 2025, Google unveiled Antigravity alongside the launch of their Gemini 3 mannequin household. Some of the shocking elements of the Antigravity launch was the return of Google co-founder Sergey Brin. After years away from day by day operations, Brin reportedly went into “Founder Mode,” working late nights alongside engineers to refine the agentic capabilities of the IDE. His purpose was to maneuver Google from “Search” to “Motion,” and Antigravity is the first instrument for that mission.

Antigravity did not seem out of skinny air. A lot of its DNA comes from a startup known as Windsurf. Google realized that the Windsurf workforce had cracked the code on how AI brokers ought to work together with an IDE, in order that they acquired the workforce in an enormous $2.4 billion deal to guide the event of what we now know as Antigravity.

Why didn’t Google simply add these options to an present instrument? As a result of conventional IDEs — like Visible Studio Code (VS Code) — had been constructed for people to sort in. They assume a human is the first actor. Antigravity is rebuilt from the bottom up, assuming an AI agent is the first employee. It’s a workspace designed for “vibe coding,” the place the velocity of your concepts is the one factor that issues, not the velocity of your typing.

 

Getting Began with Antigravity

 
Earlier than you can begin orchestrating brokers, you might want to get the software program onto your machine. As a result of Antigravity is constructed on a fork of VS Code, the set up course of feels acquainted; nonetheless, there are just a few system-specific concerns to remember to make sure the AI brokers have ample assets to run easily.

 

// System Necessities

Antigravity is not a lightweight textual content editor; it is working native fashions and browser-based brokers. Here’s what you want:

  • macOS: You want Monterey (model 12) or later. Whereas Intel Macs are supported, the expertise is optimized for Apple Silicon (M1/M2/M3/M4), as native AI inference depends closely on the unified reminiscence structure
  • Home windows: You want a 64-bit model of Home windows 10 or 11. It’s extremely advisable to put in the IDE in your major system drive (often the C: drive) to keep away from any permissions points when the AI agent tries to execute terminal instructions
  • Linux: It helps most trendy distributions like Ubuntu (20.04+), Debian (10+), and Fedora (36+). Simply guarantee you’ve got glibc 2.28 or increased

 

// The Obtain Course of

To get the reliable, newest model, skip the third-party mirrors and go on to the official supply: antigravity.google/obtain

As soon as there, you’ll see particular installers for every working system (OS). Obtain the one which matches your machine, run the installer, and launch the applying. You’ll be requested to arrange your themes and describe how you plan to make use of the Antigravity agent (please stick to the advisable review-driven growth method). Then, you’ll be prompted to “Register with Google.” Use the Gmail account you need related together with your AI quotas and cloud initiatives.

 

Sign in with GoogleSign in with Google
Signing in with Google to sync cloud initiatives | Picture by Creator

 

Navigating the Interface

 
The format is designed to maintain you from alt-tabbing between totally different home windows. Listed below are the 5 areas you’ll spend probably the most time in:

Editor Panel: The middle stage. That is the VS Code surroundings the place the agent-generated code seems. You possibly can nonetheless soar in and kind manually everytime you need.

 

Editor PanelEditor Panel
The principle editor panel in Antigravity | Picture by Creator

 

Plan vs. Quick Mode: Within the agent panel, you’ll see a toggle.

  • Plan Mode is for large duties; it generates a “Plan Artifact” so that you can assessment earlier than it begins coding
  • Quick Mode is for fast edits, like “heart this div” or “rename this variable”

 

Plan vs. Fast ModePlan vs. Fast Mode
Toggling between Plan and Quick modes | Picture by Creator

 

Agent Supervisor: That is your mission management. It lists each lively job, exhibits the agent’s “reasoning” steps, and offers standing indicators (Operating, Success, or Failed)

 

Agent ManagerAgent Manager
Managing lively AI agent duties | Picture by Creator

 

Artifacts Panel: Consider this because the “Paper Path.” It shops the particular logs, code diffs, and plans that the agent produces

 

Artifacts PanelArtifacts Panel
Reviewing generated artifacts and logs | Picture by Creator

 

Browser Preview: A built-in model of Chrome. When the agent builds an online app, it routinely launches it right here for testing

 

Browser PreviewBrowser Preview
Testing the online software within the built-in browser | Picture by Creator

 

Constructing Your First Undertaking: An Limitless Runner Sport

 
To see the facility of agent-driven growth, we may have Antigravity construct a basic infinite runner recreation from scratch utilizing HTML, CSS and JavaScript. The purpose is to construct an infinite runner internet recreation the place the participant controls a automotive touring upwards. The target is to keep away from oncoming visitors (vehicles transferring downwards). The sport options problem ranges (Straightforward, Medium, Exhausting) and growing velocity because the participant progresses.

 

// Step 1

Launch Antigravity and create a brand new, empty folder. Open this folder within the IDE.

 

// Step 2

Open the Agent Supervisor and click on New Process. Paste the next immediate into the field:

Construct an infinite runner recreation, the place there’s a automotive touring upward in an infinite working method, there are different vehicles coming down in the direction of it, and it has to keep away from them in order to not get hit, as a result of as soon as it hits it, it’s recreation over. The sport degree may be modified from simple, medium, to laborious, and the upper you go within the recreation, the quicker the vehicles come at you to hit you!

 

Project PromptProject Prompt
Getting into the venture immediate into the agent supervisor | Picture by Creator

 

// Step 3

When you press Enter, the agent initiates its workflow: it analyzes the immediate supplied and proceeds to the starting stage of the venture. As soon as that’s accomplished, coding begins instantly. You could be prompted to simply accept some opinions throughout this course of. Lastly, throughout the verification stage, the agent will truly try and run the script. If it encounters an error (equivalent to a lacking library), it can routinely set up the required library and check out once more.

Here’s what the implementation plan generated by the agent seems to be like:

 

Implementation PlanImplementation Plan
The AI-generated implementation plan | Picture by Creator

 

The walk-through file containing your entire steps adopted when constructing the infinite runner recreation:

 

Walk-through FileWalk-through File
Reviewing the step-by-step walk-through | Picture by Creator

 

Here’s what the ultimate output of the sport seems to be like:

 

Final OutputFinal Output
The ultimate recreation interface | Picture by Creator

 

You progress left and proper to keep away from the upcoming vehicles.

 

Game PlayGame Play
Taking part in the infinite runner recreation | Picture by Creator

 

In the event you hit these vehicles, it is recreation over.

 

Game OverGame Over
The sport over display screen | Picture by Creator

 

Wrapping Up

 
Google Antigravity is a glimpse right into a future the place “writing code” is only one small a part of being a developer. By delegating the implementation, testing, and debugging to Gemini-powered brokers, you might be free to give attention to the high-level structure and consumer expertise.

To get probably the most out of Antigravity, you might want to change the way you talk together with your instruments:

  • Be Particular however Purpose-Oriented: Do not inform the agent write the loop; inform it what the outcome ought to be
  • Evaluate the Plan: In “Plan Mode,” at all times learn the implementation plan earlier than clicking “Approve.” It saves time in the long term
  • Isolate Duties: Give the agent one clear mission at a time. If you wish to create a login web page and carry out a database migration, run these duties as two separate operations within the Agent Supervisor

 
 

Shittu Olumide is a software program engineer and technical author obsessed with leveraging cutting-edge applied sciences to craft compelling narratives, with a eager eye for element and a knack for simplifying advanced ideas. You too can discover Shittu on Twitter.



ICE in Minneapolis: What 3,000 ICE and CBP brokers are doing to residents

0


This story appeared in The Logoff, a every day e-newsletter that helps you keep knowledgeable in regards to the Trump administration with out letting political information take over your life. Subscribe right here.

Welcome to The Logoff: Tensions are rising in Minneapolis because the Trump administration continues its crackdown.

What’s taking place? There are some 3,000 Division of Homeland Safety brokers — each ICE and Customs and Border Safety, or CBP — in Minnesota this week, largely within the Minneapolis space. Because the killing of Renee Good by an ICE agent eight days in the past, an enormous quantity of video and reporting has documented additional brutality by federal immigration brokers, usually indiscriminate and unprovoked, towards immigrants and Americans alike.

On Wednesday night time, a federal agent shot and injured a Venezuelan man after an alleged visitors cease, giving recent gasoline to protests. And on Thursday morning, President Donald Trump threatened to invoke the Revolt Act and deploy troops to Minnesota, “if the corrupt politicians of Minnesota don’t obey the regulation and cease the skilled agitators and insurrectionists from attacking the Patriots of I.C.E.”

Why does this matter? For the second time in six years, Minnesota looks like a tinderbox. Officers within the state, together with Gov. Tim Walz and Minneapolis Mayor Jacob Frey, are urging calm and asking protesters to stay peaceable, nevertheless it more and more looks like this can be a struggle the Trump administration needs to choose. On Wednesday, senior Trump aide Stephen Miller described arresting “insurrectionists” in Minneapolis as a “nationwide safety precedence.”

What’s the context? ICE, which makes up nearly all of the brokers at present in Minneapolis, has grown considerably within the final yr, concurrently its requirements have dropped precipitously. On the similar time, underneath stress to make extra immigration arrests, they’re taking an more and more militarized strategy at odds with how ICE operated underneath earlier administrations. All of these components are on show proper now in Minneapolis.

What’s the large image? What’s taking place to Minneapolis residents already appears much less like immigration enforcement and extra like an occupation. If Trump follows by means of together with his Revolt Act risk, issues may develop far worse.

And with that, it’s time to sign off…

As at all times, thanks for studying, have a terrific night, and we’ll see you again right here tomorrow!

Beneath the ice: Satellites assist map Antarctica’s subglacial floor like by no means earlier than

0


One of many least-mapped planetary surfaces in our photo voltaic system is nearer to house than you would possibly anticipate: the continent of Antarctica.

Whereas Antarctica‘s icy floor is pretty well-studied, its subglacial bedrock panorama — situated as much as 3 miles (4.8 km) beneath the ice — is tougher to discern. Present strategies of mapping require costly ground-based and airborne surveys, and such actions are few and much between.

Cointegration or spurious regression? – The Stata Weblog

0


(newcommand{betab}{boldsymbol{beta}})Time-series information usually seem nonstationary and likewise are likely to comove. A set of nonstationary collection which can be cointegrated implies existence of a long-run equilibrium relation. If such an equlibrium doesn’t exist, then the obvious comovement is spurious and no significant interpretation ensues.

Analyzing a number of nonstationary time collection which can be cointegrated gives helpful insights about their long-run habits. Think about long- and short-term rates of interest such because the yield on a 30-year and a 3-month U.S. Treasury bond. In line with the expectations speculation, long-term rates of interest are decided by the typical of anticipated future short-term charges. This suggests that the yields on the 2 bonds can not deviate from each other over time. Thus, if the 2 yields are cointegrated, any affect to the short-term charge results in changes within the long-term rate of interest. This has necessary implications in making varied coverage or funding choices.

In a cointegration evaluation, we start by regressing a nonstationary variable on a set of different nonstationary variables. Suprisingly, in finite samples, regressing a nonstationary collection with one other arbitrary nonstationary collection often ends in important coefficients with a excessive (R^2). This provides a misunderstanding that the collection could also be cointegrated, a phenomenon generally often called spurious regression.

On this submit, I exploit simulated information to point out the asymptotic properties of an peculiar least-squares (OLS) estimator below cointegration and spurious regression. I then carry out a check for cointegration utilizing the Engle and Granger (1987) technique. These workout routines present a great first step towards understanding cointegrated processes.

Cointegration

For ease of exposition, I think about two variables, (y_t) and (x_t), which can be built-in of order 1 or I(1). They are often remodeled to a stationary or I(0) collection by taking first variations.

The variables (y_t) and (x_t) are cointegrated if their linear mixture is I(0). This suggests (betab [y_t,x_t]’ = e_t), the place (betab) is the cointegrating vector and (e_t) is a stationary equilibrium error. Usually, with the addition of extra variables, a couple of cointegrating vector might exist. Nonetheless, the Engle–Granger technique assumes a single cointegrating vector whatever the variety of variables being modeled.

A normalizing assumption reminiscent of (betab = (1,-beta)) is made to uniquely establish the cointegrating vector utilizing OLS. The sort of normalization specifies which variables happen on the left- and right-hand facet of the equation. Within the case of cointegrated variables, the selection of normalization is irrelevant asymptotically. Nonetheless, in follow, they’re sometimes guided by some concept that dictates their relationship. The normalization within the present utility results in the next regression specification.

start{equation}
label{eq1}
y_t = beta x_t + e_t
tag{1}
finish{equation}

The equation above describes the long-run relation between (y_t) and (x_t). That is additionally known as a “static” regression as a result of no different dynamics within the variables and serial correlation within the error time period are assumed.

The OLS estimator is given by (hat{beta}= frac{sum_{t=1}^T y_t x_t}{sum_{t=1}^T x^2_t}). As a result of (y_t) and (x_t) are each I(1) processes, the numerator and denominator converge to difficult capabilities of Brownian motions as (Tto infty). Nonetheless, (hat{beta}) nonetheless persistently estimates the true (beta) no matter whether or not (x_t) is correlated with (e_t). In truth, the estimator is superconsistent, which suggests it converges to the true worth at a quicker charge than an OLS estimator of a regression on stationary collection. Inference on (hat{beta}) isn’t straightforwad as a result of its asymptotic distribution is nonstandard and likewise depends upon whether or not deterministic phrases reminiscent of fixed and time pattern are specified.

Monte Carlo simulations

I carry out Monte Carlo simulations (MCS) with 1,000 replications to plot the empirical densities of the OLS estimator (hat{beta}) below spurious regression and cointegration. Within the case of spurious regression, the empirical density of (hat{beta}) doesn’t shrink towards the true worth even after rising the pattern measurement. This suggests that the OLS estimator is inconsistent. Alternatively, if the collection are cointegrated, we count on the empirical density of (hat{beta}) to converge to its true worth.

Knowledge-generating course of for spurious regression

I generate (y_t) and (x_t) in response to the next specification,

start{align}
label{eq2}
y_t &= 0.7 x_t + e_t
nonumber
x_t &= x_{t-1} + nu_{xt}
nonumber
e_t &= e_{t-1} + nu_{yt} tag{2}
finish{align}

the place (nu_{xt}) and (nu_{yt}) are IID N(0,1). (x_t) and (e_t) are unbiased random walks. This can be a spurious regression as a result of the linear mixture (y_t – 0.7x_t = e_t) is an I(1) course of.

Knowledge-generating course of for cointegration

I generate (y_t) and (x_t) in response to the next specification,

start{align}
label{eq3}
y_t &= 0.7 x_t + e_{yt}
nonumber
x_t &= x_{t-1} + e_{xt} tag{3}
finish{align}

the place (x_t) is the one I(1) course of. I enable for contemporaneous correlation in addition to serial correlation by producing the error phrases (e_{yt}) and (e_{xt}) as a vector shifting common (VMA) course of with 1 lag. The VMA mannequin is given by

start{align*}
e_{yt} &= 0.3 nu_{yt-1} + 0.4 nu_{xt-1} + nu_{yt}
e_{xt} &= 0.7 nu_{yt-1} + 0.1 nu_{xt-1} + nu_{xt}
finish{align*}

the place (nu_{yt}) and (nu_{xt}) are drawn from a standard distribution with imply 0 and variance matrix ({bf Sigma} = left[begin{matrix} 1 & 0.7 0.7 & 1.5 end{matrix}right]). The graph beneath plots the empirical densities. The code for performing the simulations is supplied within the Appendix.

Within the case of spurious regression, the OLS estimator is inconsistent as a result of it doesn’t converge to its true worth even after rising the pattern measurement from 100 to 1,000. The vertical line at 0.7 represents the true worth of (beta). Furthermore, in finite samples, the coefficients in a spurious regression are often important with a excessive (R^2). In truth, Phillips (1986) confirmed that the (t) and (F) statistics diverge in distribution as (Tto infty), thus making any inferences invalid.

Within the case of cointegration, I artificially launched serial correlation within the error phrases within the data-generating course of. This led to a biased estimate as seen within the graph above for a pattern measurement of 100, which improves barely because the pattern measurement will increase to 200. Remarkably, the OLS estimator converges to its true worth as we enhance the pattern measurement to 1,000. Whereas consistency is a crucial property, we nonetheless can not carry out inference on (hat{beta}) due to its nonstandard asymptotic distribution.

Testing for cointegration

We noticed within the earlier part that the OLS estimator is constant if the collection are cointegrated. That is true even after we fail to account for serial correlation within the error phrases. To check for cointegration, we will assemble residuals primarily based on the static regression and check for the presence of unit root. If the collection are cointegrated, the estimated residuals might be near being stationary. That is the strategy within the Engle–Granger two-step technique. Step one estimates (beta) in eqref{eq1} by OLS adopted by testing for unit root within the residuals. Checks such because the augmented Dickey–Fuller (ADF) or Phillips–Perron (PP) might be carried out on the residuals. Seek advice from an earlier submit for a dialogue of those exams.

The null and various hypotheses for testing cointegration are as follows:

[
H_o: e_t = I(1) qquad H_a: e_t = I(0)
]

The null speculation states that (e_t) is nonstationary, which suggests no cointegration between (y_t) and (x_t). The choice speculation states that (e_t) is stationary and implies cointegration.

If the true cointegrating vector (betab) is understood, then we merely plug its worth in (1) and procure the (e_t) collection. We then apply the ADF check to (e_t). The check statistic could have the usual DF distribution.

Within the case of an unknown cointegrating vector (betab), the ADF check statistic primarily based on the estimated residuals doesn’t observe the usual DF distribution below the null speculation of no cointegration (Phillips and Ouliaris 1990). Moreover, Hansen (1992) reveals that the distribution of the ADF statistic additionally depends upon whether or not (y_t) and (x_t) comprise deterministic phrases reminiscent of constants and time traits. Hamilton (1994, 766) gives vital values for making inference in such instances.

Instance

I’ve two datasets that I saved whereas performing MCS within the earlier part. spurious.dta consists of two variables, (y_t) and (x_t), generated in response to the spurious regression equations in (2). coint.dta additionally consists of (y_t) and (x_t), generated utilizing the cointegration equation in (3).

First, I check for cointegration in spurious.dta.


. use spurious, clear

. reg y x, nocons

      Supply |       SS           df       MS      Variety of obs   =       100
-------------+----------------------------------   F(1, 99)        =      6.20
       Mannequin |  63.4507871         1  63.4507871   Prob > F        =    0.0144
    Residual |  1013.32308        99  10.2355867   R-squared       =    0.0589
-------------+----------------------------------   Adj R-squared   =    0.0494
       Complete |  1076.77387       100  10.7677387   Root MSE        =    3.1993

------------------------------------------------------------------------------
           y |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
           x |  -.1257176   .0504933    -2.49   0.014    -.2259073   -.0255279
------------------------------------------------------------------------------

The coefficient on x is adverse and important. I exploit predict with the resid choice to generate the residuals and carry out an ADF check utilizing the dfuller command. I exploit the noconstant choice to supress the fixed time period within the regression and lags(2) to regulate for serial correlation. The noconstant possibility within the dfuller command implies becoming a random stroll mannequin.


. predict spurious_resd, resid

. dfuller spurious_resd, nocons lags(2)

Augmented Dickey-Fuller check for unit root         Variety of obs   =        97

                               ---------- Interpolated Dickey-Fuller ---------
                  Check         1% Crucial       5% Crucial      10% Crucial
               Statistic           Worth             Worth             Worth
------------------------------------------------------------------------------
 Z(t)             -1.599            -2.601            -1.950            -1.610

As talked about earlier, the vital values of the DF distribution is probably not used on this case. The 5% vital worth supplied within the first row of Hamilton (1994, 766) below case 1 in desk B.9 is (-)2.76. The check statistic of (-)1.60 on this case implies failure to reject the null speculation of no cointegration.

I carry out the identical two-step technique for coint.dta beneath.


. use coint, clear

. reg y x, nocons

      Supply |       SS           df       MS      Variety of obs   =       100
-------------+----------------------------------   F(1, 99)        =   3148.28
       Mannequin |  4411.48377         1  4411.48377   Prob > F        =    0.0000
    Residual |   138.72255        99  1.40123788   R-squared       =    0.9695
-------------+----------------------------------   Adj R-squared   =    0.9692
       Complete |  4550.20632       100  45.5020632   Root MSE        =    1.1837

------------------------------------------------------------------------------
           y |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
           x |   .7335899   .0130743    56.11   0.000     .7076477    .7595321
------------------------------------------------------------------------------

. predict coint_resd, resid

. dfuller coint_resd, nocons lags(2)

Augmented Dickey-Fuller check for unit root         Variety of obs   =        97

                               ---------- Interpolated Dickey-Fuller ---------
                  Check         1% Crucial       5% Crucial      10% Crucial
               Statistic           Worth             Worth             Worth
------------------------------------------------------------------------------
 Z(t)             -5.955            -2.601            -1.950            -1.610

Evaluating the DF check statistic of (-)5.95 with the suitable vital worth of (-)2.76, I reject the null speculation of no cointegration on the 5% stage.

Conclusion

On this submit, I carried out MCS to point out the consistency of the OLS estimator below cointegration. I used the Engle–Granger two-step strategy to check for cointegration in simulated information.

References

Engle, R. F., and C. W. J. Granger. 1987. Co-integration and error correction: Illustration, estimation, and testing. Econometrica 55: 251–276.

Hamilton, J. D. 1994. Time Collection Evaluation. Princeton: Princeton College Press.

Hansen, B. E. 1992. Effcient estimation and testing of cointegrating vectors within the presence of deterministic traits. Journal of Econometrics 53: 87–121.

Phillips, P. C. B. 1986. Understanding spurious regressions in econometrics. Journal of Econometrics 33: 311–340.

Phillips, P. C. B., and S. Ouliaris. 1990. Asymptotic properties of residual primarily based exams for cointegration. Econometrica 58: 165–193.

Appendix

Code for spurious regression

The next code performs an MCS for a pattern measurement of 100.

Code block 1: spurious100.do


cscript
set seed 2016

native MC = 1000

quietly postfile spurious beta_t100 utilizing t100, change

forvalues i=1/`MC' {
        quietly {
                drop _all

                set obs 100
                gen time = _n
                tsset time

                gen nu_y = rnormal(0,0.7)
                gen nu_x = rnormal(0,1.5)

                gen err_y = nu_y in 1
                gen err_x = nu_x in 1
                change err_y = l.err_y + nu_y in 2/l
                change err_x = l.err_x + nu_x in 2/l

                gen y = err_y in 1
                gen x = err_x
                change y = 0.7*x + err_y in 2/l
                if (`i'==1) save spurious, change
                qui reg y x, nocons
        }
                submit spurious (_b[x])
}
postclose spurious

Code for cointegration

The next code performs an MCS for a pattern measurement of 100.

Code block 2: coint100.do


cscript
set seed 2016

native MC = 1000

quietly postfile coint beta_t100 utilizing t100, change

forvalues i=1/`MC' {
        quietly {
                drop _all

                set obs 100
                gen time = _n
                tsset time

                matrix V = (1,0.7.7,1.5)
                drawnorm nu_y nu_x, cov(V)

                gen err_y = nu_y in 1
                gen err_x = nu_x in 1
                change err_y = 0.3*l.nu_y + 0.4*l.nu_x ///
                        + nu_y in 2/l
                change err_x = 0.7*l.nu_y + 0.1*l.nu_x ///
                        + nu_x in 2/l

                gen x = err_x in 1
                change x = l.x + err_x in 2/l
                gen y = 0.7*x + err_y
                if (`i'==1) save coint, change
                qui reg y x, nocons
        }
                submit coint (_b[x])
}
postclose coint

Combining graphs

Assuming you’ve gotten run an MCS for pattern sizes of 200 and 1,000, chances are you’ll mix the graph utilizing the next code block.

Code block 3: grcombine.do


/*Spurious regression*/
use t100, clear
quietly merge 1:1 _n utilizing t200
drop _merge
quietly merge 1:1 _n utilizing t1000
drop _merge
kdensity beta_t100, n(1000) generate(x_100 f_100) ///
        kernel(gaussian) nograph
label variable f_100 "T=100"
kdensity beta_t200, n(1000) generate(x_200 f_200) ///
        kernel(gaussian) nograph
label variable f_200 "T=200"
kdensity beta_t1000, n(1000) generate(x_1000 f_1000) ///
        kernel(gaussian) nograph
label variable f_1000 "T=1000"
graph twoway (line f_100 x_100) (line f_200 x_200) ///
        (line f_1000 x_1000), legend(rows(1))      ///
        subtitle("Spurious regression")            ///
        saving(spurious, change) xmlabel(0.7)     ///
        xline(0.7, lcolor(black)) nodraw

/*Cointegration*/
use t100, clear
quietly merge 1:1 _n utilizing t200
drop _merge
quietly merge 1:1 _n utilizing t1000
drop _merge
kdensity beta_t100, n(1000) generate(x_100 f_100) ///
        kernel(gaussian) nograph
label variable f_100 "T=100"
kdensity beta_t200, n(1000) generate(x_200 f_200) ///
        kernel(gaussian) nograph
label variable f_200 "T=200"
kdensity beta_t1000, n(1000) generate(x_1000 f_1000) ///
        kernel(gaussian) nograph
label variable f_1000 "T=1000"
graph twoway (line f_100 x_100) (line f_200 x_200) ///
        (line f_1000 x_1000), legend(rows(1))      ///
        subtitle("Cointegration")                  ///
        saving(cointegration, change)             ///
        xline(0.7, lcolor(black)) nodraw

graph mix spurious.gph cointegration.gph,      ///
      title("Empirical densities of {&beta}")



8 CIO suggestions for ERP implementation in 2026

0


Enterprise useful resource planning (ERP) suites have lengthy underpinned international enterprise operations. Their function has been to codify finest practices, optimize enterprise processes, enhance knowledge administration and assist knowledgeable decision-making. Agentic AI, nevertheless, creates a possibility to rethink how an ERP implementation helps core enterprise capabilities. 

Embedding agentic AI inside ERP methods opens the door to significant enterprise transformation by enabling new, extra environment friendly working fashions. Agentic AI can reshape how organizations work — bettering agility, effectivity and responsiveness to buyer wants. 

For organizations already invested in monolithic ERP and built-in monetary suites, prepackaged agentic capabilities present a sensible and accessible entry level. In reality, for a lot of enterprises, ERP would be the earliest and most sensible place to use agentic AI to actual enterprise issues. In an agentic group, small, multidisciplinary groups will oversee “agent factories” that handle whole processes from finish to finish. Crucially, agentic AI will rework, relatively than substitute, most ERP methods. 

“In what was as soon as a sleepy functions house, ERP is being reimagined with agentic AI capabilities,” is how John Van Decker, a distinguished analyst at Dresner Advisory Providers, defined the change. He mentioned agentic AI’s potential to prioritize duties and work together with functions throughout finance, provide chain and HR ecosystems — and to repeatedly enhance by means of suggestions and studying — transforms ERP “from a set of software program instruments into an clever workforce.”                  

Associated:What CIOs ought to find out about uncommon earths

The shift: From passive instruments to proactive brokers 

For the reason that rise of the trendy enterprise, folks have managed core enterprise processes throughout finance, provide chain and manufacturing. Administration theorist Gary Hamel has noticed that most of the processes nonetheless in use at the moment had been initially designed in the course of the early Industrial Revolution. Now, many duties that people have carried out for greater than a century — and which had been later embedded inside ERP implementations — are shifting to agentic AI. Right here, brokers will more and more tackle work throughout finance and planning, provide chain administration and procurement, manufacturing, HR and customer support.

Agentic AI gives the potential to evolve from passive to proactive administration, enabling impartial multi-step duties and choices. Autonomous “digital employees” will handle and execute subtle, end-to-end enterprise processes. Van Decker calls out the next areas of affect: 

Associated:As SaaS spend grows, CIOs concentrate on software sprawl

  • Finance and planning. Brokers will automate features together with bill processing, expense reporting, monetary assertion technology and compliance verification. Moreover, they are going to independently detect anomalies and supply liquidity forecasts.

  • Provide chain administration and procurement. Brokers will monitor stock ranges, renegotiate provider contracts in actual time in response to commodity worth fluctuations, modify manufacturing schedules to accommodate materials delays and optimize logistics networks.

  • Manufacturing. Brokers will improve operational effectivity by means of real-time monitoring, predictive upkeep scheduling, automated high quality assurance powered by pc imaginative and prescient, and adaptive reconfiguration of manufacturing unit flooring processes to deal with disruptions.

  • HR. Brokers will streamline recruiting processes, handle onboarding documentation, deal with routine payroll inquiries and reply to worker advantages questions, thereby permitting HR professionals to concentrate on strategic initiatives.

  • Customer support. Brokers will handle Degree 1 assist interactions, carry out sentiment evaluation, entry order histories, ship customized communications and escalate complicated points to human representatives with complete context.

2026 technique: Priorities for CIOs and CFOs

Given these alternatives, what actions ought to CIOs — and the CFOs they assist — take? The next priorities must be added to the CIO agenda for 2026:

  1. Develop a transparent marketing strategy for adopting agentic AI, together with its implications for the present ERP surroundings, working mannequin, and monetary controls.

  2. Have interaction ERP distributors early to know their agentic AI roadmaps, timelines for adoption and the operational, safety and monetary dangers these capabilities might introduce.

  3. Outline how people and agentic methods will work collectively, establishing express “human-in-the-loop” roles, choice rights and escalation paths. AI is highly effective — however not infallible.

  4. Set up a complete threat mitigation framework that addresses safety threats (e.g., immediate injection), moral dangers (corresponding to biased choices), infrastructure and knowledge high quality points, implementation challenges and organizational boundaries, together with weak governance or unrealistic expectations.

  5. Handle foundational readiness gaps, together with immature instruments, inconsistent knowledge high quality, unclear processes and the problem of measuring ROI from agentic AI initiatives.

  6. Embed governance, transparency, safety and vendor lock-in issues into each agentic AI initiative from the outset, relatively than treating them as afterthoughts.

  7. Plan for the specialised expertise required to design, deploy and handle agentic AI methods, recognizing that efficient implementation will depend on scarce and evolving experience.

  8. Construct a proactive communication and alter administration technique to deal with cultural resistance, worker considerations about job affect and management warning round threat and value. Clear intent and considerate rollout shall be essential to adoption at scale.

To optimize or rework: That’s the query

ERP ought to proceed to be a core factor of each CIO’s 2026 agenda. For main organizations, the central problem shall be deciding whether or not to deploy brokers to optimize present enterprise processes or to pursue game-changing enterprise transformation. Because of this, AI-savvy CIOs will work to strike the suitable steadiness between incremental course of enchancment and cost-cutting and daring reinvention. Our knowledge exhibits that organizations with mature, industrialized knowledge and course of foundations are already taking a broader, extra transformative strategy to agentic AI.



The Obtain: spying on the spies, and promising local weather tech


That is as we speak’s version of The Obtain, our weekday e-newsletter that gives a each day dose of what’s occurring on this planet of expertise.

Meet the person searching the spies in your smartphone

In April 2025, Ronald Deibert left all digital units at house in Toronto and boarded a aircraft. When he landed in Illinois, he purchased a brand new laptop computer and iPhone. He wished to cut back the chance of getting his private units confiscated, as a result of he knew his work made him a major goal for surveillance. “I’m touring underneath the idea that I’m being watched, proper down to precisely the place I’m at any second,” Deibert says. 

Deibert directs the Citizen Lab, a analysis middle he based in 2001 to function “counterintelligence for civil society.” Housed on the College of Toronto, it’s one of many few establishments that examine cyberthreats completely within the public curiosity, and in doing so, it has uncovered a number of the most egregious digital abuses of the previous twenty years.

For a few years, Deibert and his colleagues have held up the US as the usual for liberal democracy. However that’s altering. Learn the total story.

—Finian Hazen

This story is from the newest challenge of our print journal. If you happen to subscribe now to obtain future copies after they land you’ll profit from some massive reductions, and get a free tote bag! 

Three local weather applied sciences breaking by way of in 2026  

—Casey Crownhart 

Pleased New Yr! I do know it’s a bit late to say, however it by no means fairly feels just like the yr has began till the brand new version of our 10 Breakthrough Applied sciences checklist comes out. 

For 25 years, MIT Know-how Assessment has put collectively this package deal, which highlights the applied sciences that we expect are going to matter sooner or later. This yr’s model has a bunch of local weather and vitality picks together with sodium-ion batteries, next-generation nuclear, and hyperscale AI information facilities. Let’s check out what ended up on the checklist, and what it says about this second for local weather tech. 

This story ran in The Spark, our weekly e-newsletter all concerning the applied sciences we will use to fight local weather change. Join to get it in your inbox first each Wednesday. 

And, if you happen to’re eager to study extra about why AI firms are betting massive on next-gen nuclear, be part of us for an unique subscriber-only Roundtable occasion on Wednesday January 28 at 2pm ET. 

The must-reads

I’ve combed the web to seek out you as we speak’s most enjoyable/essential/scary/fascinating tales about expertise.

1 AI firms are actually deeply entwined with the US army
And it seems to be like they’re solely set to get nearer. (Wired $)
Three open questions concerning the Pentagon’s push for generative AI. (MIT Know-how Assessment)

2 Grok will adjust to native legal guidelines, X has stated
A worldwide backlash over customers creating ‘undressing’ photographs of actual folks appears to have compelled its hand. (BBC)
+ To this point there’s no proof it’s truly following by way of on that promise although. (The Verge)
Elon Musk may cease all of it immediately if he                         wished to. (Engadget)

3 The dangers of utilizing AI in faculties outweigh the advantages
Based on a sweeping new research by the Brookings Establishment’s Middle for Common Training. (NPR)
AI’s giants try to take over the classroom. (MIT Know-how Assessment)  

4 Trump is imposing new tariffs on high-end chips
They’re fairly slender although, and depart loads of room for exports to China. (WP $)
Zhipu AI says it’s skilled its first main mannequin solely on Chinese language chips. (South China Morning Publish)

5 A UK police power blamed Microsoft Copilot for an intelligence error 
After spending weeks denying it was utilizing AI instruments in any respect. (Ars Technica)
Nervous about police and legal professionals utilizing AI? Nicely, judges are at it too. (MIT Know-how Assessment

6 Contained in the compounds the place the fraud business makes its billions
The main points are grim—for instance the very fact staff struck a gong each time they scammed somebody out of $5,000. (NYT $)
+ Inside a romance rip-off compound—and the way folks get tricked into being there. (MIT Know-how Assessment)

7 Bandcamp has banned purely AI-generated music from its platform 
It’s the primary on-line music platform to take this step. (Billboard)
Can AI generate new concepts? (NYT $)

8 Keep in mind Havana Syndrome? The US could have discovered the system that causes it
It was acquired for tens of millions of {dollars} underneath the final administration, and it’s nonetheless being studied. (CNN)

9 This research didn’t show social media time causes teenagers’ psychological well being points
It’s a typical assumption, however there’s nonetheless remarkably little proof to again it up. (The Guardian)

10 The UK is planning to construct a record-breaking variety of wind farms
Its authorities is pushing for the overwhelming majority of the nation’s electrical energy to return from clear sources by 2030. (BBC)

Quote of the day

“Ladies and ladies are way more reluctant to make use of AI. This ought to be no shock to any of us. Ladies don’t see this as thrilling new expertise, however as merely new methods to harass and abuse us and attempt to push us offline.”

—Clare McGlynn, a legislation professor at Durham College, tells The Guardian she fears that using AI to hurt ladies and ladies is just going to develop. 

Another factor

DANIEL STOLLE

Contained in the little-known group setting the company local weather agenda

As 1000’s of firms trumpet their plans to chop carbon air pollution, a small group of sustainability consultants has emerged because the go-to arbiter of company local weather motion.

The Science Primarily based Targets initiative, or SBTi, helps companies develop a timetable for motion to shrink their local weather footprint by way of some mixture of chopping greenhouse-gas air pollution and eradicating carbon dioxide from the environment. After years of small-scale sustainability work, SBTi is rising quickly, and governments are paying consideration. 

However whereas the group has earned reward for reeling the personal sector into constructive conversations about local weather emissions, its rising affect has additionally attracted scrutiny and raised questions on why a single group is setting the requirements for most of the world’s largest firms. Learn the total story.

—Ian Morse

We are able to nonetheless have good issues

A spot for consolation, enjoyable and distraction to brighten up your day. (Acquired any concepts? Drop me a line or skeet ’em at me.)

+ The leaders of Japan and South Korea drummed up a viral second with a jam session this week. 
+ Wrestle throughout the chilly, darkish winter months? Right here’s how you can make issues simpler for your self
+ If you happen to like getting misplaced within the depths of Wikipedia, Freakpages is for you. 
+ From Pluribus to Stranger Issues, we actually can’t get sufficient of hive mindsin tales currently. ($)

Cyber crime is up — fortify your Mac for 44% off with Moonlock

0


TL;DR: Get cyber safety particularly designed to your Mac with Moonlock, an app that makes antivirus easy. And proper now, you will get a yr’s subscription for simply $29.99 (Reg. $54.00).

Cybercrime is up. The stats present that, because of AI and an more and more subtle menace panorama, even your Mac could be prone to contracting malware and being accessed by dangerous actors. However with Moonlock, you get personalised safety recommendation and top-of-the-line prevention.

Lastly take your Mac safety significantly with Moonlock

For many years, MacBooks have been recognized to be nearly impenetrable. Nevertheless, even your Mac is in danger on this scary menace panorama. However that does not imply your built-in Mac software program is instantly ineffective. It is an incredible first line of protection. This cybersecurity app takes your safety to the following degree.

Here is the factor: cybersecurity is not one thing most of us need to cope with. We simply wish to know that our units are secure from bots and hackers. Moonlock provides a plug-and-play resolution that enhances your Mac’s security measures to maintain you secure from Mac-specific malware and provide the actual safety frameworks you want to your gadget and your life.

Remove threat for an reasonably priced value

Again within the day, viruses have been straightforward to detect. Now, they’re hidden and much more harmful. Moonlock retains you protected at an incredible value, so you do not have to sacrifice security for affordability.

For only one fee, Moonlock provides options like:

  • Quarantine management: Disables any detected malware so you may resolve your subsequent transfer.
  • Constructed-in VPN: No must pay for a VPN. You get safety and privateness in a single app.
  • Intuitive interface: A human-friendly dashboard that you do not want tech experience to work.
  • Superior menace discovery: confirmed effectiveness at discovering hidden threats.

Do not miss your likelihood to get 44% off a yr of Moonlock. Obtain the app now for simply $29.99 (Reg. $54.00).

Moonlock: Cyber Safety for Mac

See Deal

StackSocial costs topic to alter.

See different gadgets within the store.



The most effective folding electrical bikes for 2026, examined and reviewed

0


All types of fascinating one- and two-wheeled automobiles crisscross my metropolis streets, together with an rising variety of reasonably priced, moveable types of tech-packed private transportation . Spandex-clad cyclists make the most of protected lanes, whereas electrical scooters zip over sidewalks (and anyplace else). Sometimes, an odd unicyclist or one-wheeler breaks up the predictability. I significantly like to look at foldable e-bike riders who go from path to coach, remodeling their private transit right into a dice to hold by means of a crowd. These now-you-see-it, now-you-don’t collapsible electrical bikes provide energy help for commuting or errands with out taking over priceless actual property in an condo or townhouse. These versatile choices can typically match into an SUV or RV for prolonged out of doors adventures with out the necessity for a heavy-duty, and sometimes costly, rack. We’ve up to date probably the most dependable choices—like our greatest general, the Lectric XP4—so you could find the most effective folding electrical bikes to match your driving type and desires.

How we chosen the most effective folding electrical bikes

We don’t check these bikes in a lab—we take them out into the wild. Which means folding and unfolding them on sidewalks, rolling them into elevators to see if they really match, and lifting them into automobile trunks to search out out if it’s a one-person job or a two-person battle. We trip the identical loops we’d take to work or the shop—full with potholes, stoplights, sketchy shoulders, and loads of zippy scooters.

Each bike will get pushed on efficiency: full pedal help, lights on, up and down hills till the battery faucets out. We take note of what it’s prefer to dwell with the factor—the way it rides, the way it folds, and whether or not we’d truly need to use it day by day.

The most effective folding electrical bikes: Critiques & Suggestions

Folding e-bikes are all about house effectivity, however what that appears like will depend on the rider and what equipment you may need to carry. When you’re an city commuter navigating a cramped condo, crowded prepare platforms, and 0 bike parking, you need one thing that folds quick, rolls simple, and doesn’t make enemies within the elevator. On the flip facet, RVers, van-lifers, or campers could need a extra rugged trip with sufficient energy and vary to cruise trails, run errands, or sub in for a automobile once you’re off-grid.

All folding e-bikes promise one factor: they take up much less house once you’re not driving them. Most have shorter wheelbases, decrease standover heights because of smaller diameter wheels, and upright driving positions that preserve you seen in visitors and balanced at low speeds. However the true magic is within the fold, whether or not that’s a single hinge or a multi-jointed origami routine. When you’re simply tucking it right into a hallway or closet, weight won’t be a dealbreaker. However when you’re hauling it right into a trunk or up a flight of stairs, be certain that it’s one thing you may elevate while not having an ice pack after.

When you’re used to driving full-sized bikes, folding e-bikes do take some getting used to (as do all e-bikes). Taller riders may really feel like they’re perched on a circus act, whereas smaller riders typically discover the compact match extra confidence-inspiring. The sharper turning radius is a win in tight areas, however some gearing setups can go away you spinning with out a lot payoff. Journey a number of miles, although, and most of the people settle into the rhythm rapidly.

Finest general: Lectric XP4


See It

Professionals

  • Entrance oil suspension fork
  • Rear rack that may maintain as much as 150 kilos
  • 5 ranges of pedal help plus throttle
  • Entrance and rear 602 hydraulic brakes
  • Built-in headlights and hind lights
  • Entrance and rear fenders
  • Detachable battery
  • Exterior charging port
  • Slime-lined tires to stop flats
  • A number of mounting factors for extra equipment
  • 1-year guarantee

Cons

  • Lengthy-range battery obtainable for further
  • Effectively-organized, however exterior wires and cords
  • Nothing to safe the bike when folded

Specs


  • Unfolded bike:


    66 inches lengthy by 18.5 inches broad by 45 inches tall


  • Folded bike:


    37 inches by 17 inches broad by 29 inches tall


  • Motor:


    500W hub motor with 55Nm of torque (750W possibility obtainable)


  • Battery:


    672Wh lithium-ion


  • Using vary:


    As much as 50 miles utilizing pedal help or as much as 20 utilizing throttle solely (Lengthy-range 85-mile model obtainable)


  • Class:


    1, 2, or 3


  • Prime pace:


    28 mph


  • Bike weight:


    62 kilos with out battery (7 lbs.)


  • Carrying capability:


    330 kilos (combining rider and passenger/cargo)

Lectric is without doubt one of the OGs of foldable electrical bikes, a lot in order that I’d practically all the time get a shout or recognition from one other Lectric rider after I was driving one round city. They’d chat with me at a cease or trip alongside me for a bit, and I can truthfully say it’s by no means occurred to me on one other model of e-bike. The Lectric XP4 finetunes what it’s realized from earlier fashions for an reasonably priced, enjoyable trip with 5 pedal-assist ranges and a thumb throttle. Lectric added extra juice to the 500W motor, which may peak at 1,092W and kick out 55Nm of torque.

The tires are fat-tire lite—20 by 3 inches with customized tread—which makes for a extra cushioned trip with the entrance suspension fork. The improved Shimano Altus gearing and an in-house-designed torque sensor imply much less livid pedaling than different fashions to stand up to high speeds. Lectric lists the usual step-over at $1,300 (there are additionally step-thru and long-range variations), however you may typically discover it on sale for a grand—and but, it nonetheless has hydraulic brakes, that are a rarity at this value level. The XP4 additionally comes inventory with many good commuter extras, like built-in lights, entrance and rear fenders, and a monster rear rack that may maintain as much as 150 kilos. Lectric presents a ton of equipment, together with a passenger seat for gentle riders (e.g., children). It now encompasses a TFT LCD shade show with a USB-C charging port.

For most individuals, the Lectric will meet their folding e-bike wants, but it surely’s not excellent. Lectric, greater than many manufacturers, makes you effectively conscious of cords. Whereas the wiring up entrance is well-organized, nothing is internally threaded. It folds at two factors—at mid-frame and the handlebars—and it frankly takes some apply to align pedals, handlebars, and wheels excellent. You’ll additionally want a bungee wire (or one thing related) to maintain every little thing good and tight if you wish to transfer the bike whereas folded (see under). But it surely does get sufficiently small to enter a typical automobile trunk—not a Miata, let’s not get loopy. It’s, nevertheless, a heavy 62 kilos (69 until you scrap the battery).

Lectric

Finest for city and multimodal commuters: Brompton Electrical G Line


See It

Professionals

  • One of the compact folding bikes
  • Geared up for commuting with built-in entrance and hind lights, fenders with mudflaps, rear rack, and a bell
  • Stability and dealing with like a full-size bike
  • Detachable battery pack for straightforward charging or safe storage
  • Wonderful for multi-modal commuters
  • Is available in three body sizes to suit riders from 5’ to six’6”
  • Out there in three colours: journey orange, forest inexperienced, traildust white (proven)
  • Optionally available app

Cons

  • Rocker-style show controls for PAS aren’t instantly intuitive
  • Bigger wheels imply folding takes extra apply
  • No throttle and no suspension, so plan to pedal and anticipate vibration on rougher roads
  • Expensive for informal commuters

Specs


  • Folded dimension:


    28.3 inches deep × 26.2 inches tall × 16 inches broad


  • Motor:


    250W rear-hub motor


  • Battery:


    345Wh capability, costs to 100% in 4-5 hours


  • Removeable battery:


    Sure (housed in front-mounted bag)


  • Compliance:


    UL 2849 licensed


  • Using vary:


    20 to 45 miles


  • Class:


    1


  • Prime assisted pace:


    20 mph


  • Throttle:


    No


  • Drivetrain:


    4-speed


  • Brakes:


    Tektro hydraulic disc


  • Tires:


    20-inch Schwalbe G-One Allround 54-406 tan wall 


  • Bike weight:


    About 43 kilos w/ battery bag (35 lbs. bike solely)


  • Connectivity:


    Brompton Electrical app (iOS and Android)

The Brompton Electrical G Line doesn’t appear like your common e-bike—and that’s the purpose. That is the electrified model of Brompton’s iconic folding bike, designed for individuals who want a compact trip that’s simple to stash indoors or carry onto public transit. It folds down smaller than anything on this roundup, and hides its motor and battery inside a intelligent front-mounted bag that clicks on and off the body. It’s a refined answer for riders who dwell in flats, juggle multi-leg commutes, or simply don’t have a secure spot to lock up a full-sized bike.

At about 43 kilos, the Electrical G Line is heavy for Brompton however very gentle for an e-bike, particularly a folding one. Its signature tri-fold is among the many quickest folding bikes: pivot the rear wheel beneath, collapse the principle body, and fold down the handlebars. The entire course of solely takes a few minutes, and the rear rack has wheels that allow you to roll the folded bike slightly than lug it round. Expert unfolders can kinda flick the bike open, however my quick stature by no means bought the cling of it—or it might be the G Line’s bigger 20-inch wheels. Nonetheless, for individuals who go from bike trip to subway, the entire package deal rolls simply by means of a station and may slide beneath a desk when you get to an workplace. 

Regardless of its small dimension and quirky appears to be like, the G Line rides like a a lot greater bike. The 250W rear hub motor is quiet and natural-feeling, giving a clean help that enhances your pedaling with out ever lurching ahead. It’s paired with a 345Wh battery and a four-speed drivetrain that shifts cleanly—even when you may want for an additional gear when actually pushing. The high-volume Schwalbe tires absorb most highway chatter, and the stretched body geometry provides it a secure, planted really feel. On paved streets and clean trails, it feels assured and composed. Gravel is extra of a backup plan; the G Line can deal with gentle filth or hardpack, however with out suspension, rougher terrain sends vibration straight to your fingers.

The cockpit is minimal, and that may be a blessing or a quirk relying in your type. The small shade show appears to be like nice, and you alter pedal help degree by rocking the display up or down. Nevertheless, so many management models use bodily buttons or touchscreens that this motion isn’t essentially intuitive. You can too change help ranges from the battery bag or through the Brompton app, however neither is a quick mid-ride adjustment. On a full battery with max help, it delivered over 27 miles of metropolis and path driving in testing, and it steps down its assist steadily because the battery drains. Even absolutely unpowered, the bike is straightforward to trip because of its well-balanced body and drivetrain.

The G Line isn’t an off-the-cuff buy—it begins round $4,950—but it surely’s not constructed for informal wants. It’s a splurge for riders who need the comfort of a folding bike with the trip really feel of one thing a lot bigger, and who plan to make use of it as a part of on a regular basis life. In case your bike lives inside with you, travels on trains, or will get folded twice a day, this one earns its preserve. As we defined in our full evaluate, the G Line makes few compromises for one thing this compact, and nonetheless manages to be enjoyable, purposeful, and unexpectedly quick.

A Traildust White Brompton Electric G Line folding ebike leaning against a wooded guardrail on the C&O Towpath, green grass and trees in the background
Heather Kuldell-Ware

Finest for off-road adventures: HeyBike Mars 3.0


See It

Professionals

  • 750 W (1,400 W peak) motor with critical torque
  • Full suspension and fats tires for consolation and traction
  • A number of safety layers: NFC playing cards, PIN, app, and key
  • Excessive-capacity battery for longer ranges
  • Comes with built-in digital horn, brake lights, and switch alerts
  • Is available in three colours: orange, blue and black

Cons

  • Stiff heart hinge could make eradicating the battery or folding a two-person affair
  • Dynamic battery gauge will be arduous to foretell
  • Laborious to safe to a motorcycle rack with no lengthy, versatile chain
  • It folds, but it surely nonetheless requires bigger automobiles for transport
  • It’s simple to overlook the important thing required for driving

Specs


  • Folded dimension


    38.2 inches tall x 19.7 inches broad x 33.1 inches deep


  • Motor


    750 W (1,400 W peak) rear hub with 95 Nm torque


  • Battery


    624 Wh


  • Detachable


    Sure, lockable


  • Compliance


    UL 2849


  • Vary


    As much as 65 miles


  • Class


    Ships as Class 2, will be modified to 1 or 3


  • Prime pace


    Unlock as much as 30 mph


  • Throttle


    Sure


  • Tires


    20″ × 4″ fats tires


  • Suspension


    Entrance hydraulic fork with 65mm journey and rear Horst-link


  • Drivetrain


    Shimano Altus 8-speed


  • Weight


    about 70 lb


  • Body dimension


    One, matches riders from 4’11” to six’2”


  • Connectivity


    HeyBike app

The Heybike Mars 3.0 is constructed for weekend wanderers, RV adventurers, and anybody who’d slightly be bouncing down a path than weaving by means of visitors. With 4-inch fats tires and full suspension, it turns cracked pavement and chunky gravel into one thing nearer to a suggestion than a problem. It’s much less about becoming into tight areas and extra about blasting out into wide-open ones.

Twist the throttle or faucet the pedals, and the Mars 3.0 surges ahead prefer it’s been ready for a greenlight at a race observe. In Eco mode, pedaling feels easy, and hills virtually vanish. Kick it into Enhance, and pedaling turns into non-compulsory. The plush suspension softens each hit, giving even tough trails a floaty, playful vibe. It’s the type of trip that makes you search out the bumpy route only for the enjoyable of it.

Sure, it folds, however you’ll have to imply it. At round 70 kilos, the Mars 3.0 is hefty, and the folding course of—collapsing the body and dropping the handlebars—is made trickier by its bulk and a middle hinge that started off fairly stiff. There’s a built-in stand to maintain stress off the drivetrain, however lifting it right into a automobile is a two-person job for most individuals. It slot in a hatchback, barely, and can be extra at residence rolling into an RV or the again of a pickup.

You’ll additionally have to preserve observe of the required key, and in reality, the Mars 3.0 encompasses a multilayered safety system that permits you to select between unlocking the bike with the app, PIN code, or NFC card. It appears like overkill till you keep in mind this bike is principally a joyride machine, and Heybike clearly doesn’t need it rolling away with out you.

When you consider folding as a approach to deliver your bike to the path—not essentially retailer it beneath your desk—this one makes a robust case. As we defined in our full evaluate, the Mars 3.0 trades sleekness for functionality, and for the correct of rider, that’s precisely the purpose.

A Vivid ORange Hetbike Mars 3.0 sitting on green grass
Heather Kuldell-Ware

Finest for long-range rides: Velotric Fold 1 Plus


See It

Professionals

  • Choice to modify between torque and cadence sensors
  • Pace customization obtainable
  • Totally outfitted for commuting with fenders, rear rack, 130 lux entrance gentle with adjustable angle, and rear brake gentle with flip alerts
  • Comfortable trip with entrance suspension and comfortable saddle
  • Lengthy-range and highly effective 750W motor
  • Telescoping handlebars
  • Easy, firm fold with stand to maintain from placing weight on delicate parts
  • USB-C plug for telephone charging
  • Built-in Apple Discover My
  • Velotric app connectivity
  • Is available in 4 colours: Mango, Electrical Blue (the mannequin examined), Stone Grey, and Pearl White

Cons

  • Might not match simply in customary automobile trunks 
  • Battery is usually a little tough to slip out and in of the body
  • Folding pedals really feel unintuitive at first
  • Motor will get noticeably louder on greater mode and help combos
  • Not for subway/prepare commuters

Specs


  • Folded bike:


    37.8 inches lengthy x 19.7 inches broad x 33.5 inches excessive


  • Motor:


    750W (as much as 1,100 W peak) rear hub


  • Battery:


    624Wh, IPX7 and UL 2271 Licensed


  • Using vary:


    As much as 68 miles (pedal help), as much as 45 miles (throttle solely)


  • Class:


    1, 2, or 3 (adjustable)


  • Prime pace:


    28 mph (adjustable)


  • Bike weight:


    63 kilos


  • Carrying capability:


    450 kilos complete (as much as 120 kilos on rear rack)

The Velotric Fold 1 Plus solutions the query, “Can a motorcycle be chill and nonetheless haul?” with a assured yep. Its 750W rear hub motor (1,100W peak) rockets you off the road, whereas a 624Wh battery delivers an eyebrow-raising 68 miles on pedal help. That’s “all-day journey” territory—with out the sore legs.

One dimension actually does match all right here (as our full evaluate reveals): the ultra-low step-through body welcomes riders from 4’9” to six’5” and retains cargo (as much as 120 kilos on the rear rack) secure and regular. The trip is pure cush—entrance suspension, plush saddle, upright stance, and chunky 20×3-inch puncture-resistant tires absorb bumps so you may deal with grinning.

Commuter cred? Totally earned. Full fenders, a 130-lux aimable headlight, brake-activated rear gentle with flip alerts, hydraulic disc brakes, and even a USB-C port to juice your telephone mid-ride. It folds to 37.8 x 19.7 x 33.5 inches—compact sufficient for SUVs and RVs—whereas a built-in stand and velcro strap preserve the package deal neat. Although it might be cumbersome when you’re making an attempt to go from path to coach.

Right here’s the kicker: you may tune it to your temper. Swap between torque and cadence sensors, tweak high pace as much as 28 mph (or dial it down for the children), and observe each trip by means of the Velotric app—with Apple Discover My as your digital security internet.

Quick, cozy, and ridiculously versatile, the Fold 1 Plus isn’t only a folding bike—it’s your ticket to go farther, sooner, and have extra enjoyable doing it. A do-it-all folder for riders of virtually any dimension, this e-bike goals to (and largely does) match practically everybody’s wants.

Side closeup of a blue Velotric Fold 1 Plus folding ebike completely folded in front of a grey stone wall
Heather Kuldell

Finest light-weight: Urtopia Carbon Fold 1


See It

Professionals

  • Carbon fiber body means exceptionally light-weight
  • No meeting required
  • Colourful, crisp show
  • Quick wheelbase means nimble dealing with
  • Good giant chainring means it doesn’t really feel such as you’re peddling furiously
  • Hydraulic disc brakes
  • Built-in headlight/tailight
  • Entrance and rear fenders
  • Detachable battery clearly positioned in seat put up
  • Very quick two-step fold, with magnetic fasteners to maintain issues compact throughout transport
  • Latches really feel inflexible and safe when in play

Cons

  • Carbon fiber isn’t low-cost
  • Max vary could also be extra optimistic than lifelike
  • Carbon fiber absorbs a good quantity of highway, however it may possibly’t change the consolation of precise suspension

Specs


  • Folded bike:


    800 mm x 680mm


  • Motor:


    500W peak rear hub motor w/ 42Nm torque


  • Battery:


    252Wh lithium-ion


  • Using vary:


    As much as 40 miles


  • Class:


    2


  • Prime pace:


    20 mph


  • Bike weight:


    29 kilos


  • Most rider weight:


    220 kilos

The Urtopia Carbon Fold 1 is an incredibly pleasant however diminutive Shimano Altus 8-speed commuter that folds up quick and rides even sooner. Due to its brightly painted carbon fiber body and fork (the corporate’s signature materials, Saffron Yellow colorway proven above), this bike weighs simply 29 kilos. That makes it gentle sufficient to toss into small automobile trunks or an RV for highway journeys, tenting, or different overlanding adventures. And makes it a dream when you dwell in a multi-story walkup and/or have to issue subway rides into your commute. Regardless of its compact construct (which arrives absolutely assembled), it may possibly assist riders from 5’1’’ to six’1’’ and as much as 220 kilos in complete weight, although taller or long-legged riders could really feel considerably cramped. 

Little particulars make a giant distinction: Constructed-in magnets preserve the bike securely folded when you transfer it round—no awkward flopping. Heavy-duty latches lock it again into driving place. Cable administration is clear. A central cutout serves as a deal with and place to string your chain, so nobody can fold the bike off its lock. 

Out on the highway, the 500W peak rear hub motor and 42Nm of torque present fast, zippy acceleration, permitting you to achieve speeds of as much as 20mph. Its low standover top makes it super-easy for frequent stopping and standing throughout metropolis rides. It additionally has a brief wheelbase, making it excessively nimble—nice for dodging automobiles illegally parked within the bike lane and weaving by means of potholes, cones, or individuals gazing their telephones. Plus, the TEKTRO hydraulic disc brakes are responsive. It’s surprisingly enjoyable for one thing whose foremost perform is to be sensible. 

The 252Wh battery is cleverly hidden within the seat put up (which will be absolutely eliminated for charging/storing … or to discourage thieves, as a result of a motorcycle with no seat is simply a particularly inconvenient scooter). It powers the motor, but in addition the colour display and built-in headlight/taillight. It feels prefer it’s designed for a number of quick jaunts slightly than vital mileage; nevertheless, because the revealed 40-mile vary appears optimistic, particularly on help ranges above Eco (Touring mode felt nearer to the 20s). 

My model two wishlist features a extra delicate torque sensor and an even bigger battery, however for anybody quick on house—or anybody who desires a journey bike with out messing with a rack—the Carbon Fold is a blast. The Carbon Fold 1’s really helpful retail value is $2,500, but it surely has been persistently on sale for $1,600. 

Two people carrying Urtopia Carbon Fold bikes up stairs
Urtopia

Finest worth: Ride1Up Portola


See It

Professionals

  • 5 pedal help modes plus throttle
  • Entrance suspension fork
  • Twin-piston hydraulic brakes
  • Back and front fenders
  • Rear rack holds as much as 130 kilos
  • Built-in entrance and rear bikes
  • Folding pedals
  • Chainguard keep
  • Quick velcro strap to carry the bike collectively when folded
  • 1-year guarantee

Cons

  • Lengthy-range battery obtainable for further
  • Effectively-organized, however exterior wires and cords
  • Taller riders could discover leg extension missing

Specs


  • Unfolded bike:


    66.2 inches lengthy by 19 inches broad by 44.4 inches tall


  • Folded bike:


    33 inches lengthy by 19 inches broad by 29.5 inches tall


  • Motor:


    750W geared hub motor with 65Nm torque


  • Battery:


    500Wh


  • Using vary:


    As much as 40 miles


  • Class:


    Switchable to 1, 2 or 3


  • Prime pace:


    28 mph


  • Bike weight:


    59 kilos


  • Carrying capability:


    300 kilos

The Ride1Up Portola has a giant ol’ motor—750W, the biggest of this roundup. It additionally encompasses a entrance suspension fork, 8-speed drivetrain, hydraulic brakes, and a muscular welded rear rack that may maintain as much as 130 kilos. These particulars are shocking to search out on a folding e-bike and downright surprising to search out on one which prices lower than a grand.

The Portola presents loads to love, particularly on the value level. Sturdy 20-inch by 3-inch tires and a entrance suspension fork work collectively to maintain the rider comparatively comfy. There’s just one body dimension, which Ride1Up calls “one dimension matches most,” or on this case, means riders from 4’10” to six’4”—although taller riders or long-legged ones notice they’d like slightly extra extension once they pedal. 

The Portola initially begins with a ten.4Ah battery, however an improve to 13.4Ah is just $100 extra. With 5 pedal-assist ranges and a throttle, you may want the additional juice. The vary is as much as 40—or 45 with the bigger battery. The e-bike additionally will be switched between Class 1, 2, or 3, relying on whether or not you need to restrict or unleash a possible high pace of 28 mph. And right here’s one quite simple however very good contact that many folding bike makers overlook: The Portola has a small velcro strap to safe the bike when it’s folded. It takes some fussing to get the wheels and handlebars excellent, however folding bike house owners often add their very own bungees or straps as a result of nothing was included. 

What to contemplate when trying to find the most effective folding electrical bikes

Normally, a seek for an e-bike begins with sticker shock, particularly when you don’t trip different bicycles. Yow will discover finances electrical bikes beneath $1,000, although fashions rapidly get into the a number of hundreds. Anticipate finding extra highly effective motors with extra torque, bigger batteries, and frames made from higher-quality, typically lighter-weight supplies as the worth climbs. They’re like automobiles: Yow will discover a fairly priced, dependable mannequin or spend critical bucks on high-end parts, luxurious options, and eye-grabbing designs. Whereas the choices could also be daunting, there’s a mannequin on the market to fit your tastes and finances, whether or not you’re in search of the greatest electrical commuter bike or the greatest fats tire electrical bike

It’s additionally value checking whether or not your state presents tax credit or rebates for e-bikes to encourage adoption.

E-bike lessons: What they imply for folding bikes

E-bike lessons assist outline how briskly your bike can go together with motor help and the place you’re legally allowed to trip. This issues loads when you plan to trip on bike paths, public trails, or shared-use routes. Nevertheless, the principles will range from state to state and from locality to locality. Earlier than shopping for, examine your native legal guidelines as a result of some states deal with e-bikes like common bikes, whereas others (like Alaska and New Mexico) regulate them extra like mopeds. (Try the non-profit advocacy group Folks for Bikes’ information for state legal guidelines.)

To make this even muddier, many e-bikes permit the house owners to vary lessons by means of the show or an app. That is normally known as “unlocking,” but it surely’s a easy setting that permits you to set a high pace of 20 or 28 miles per hour. Equally, some e-bikes have throttles that may be enabled, disabled, or completely eliminated. 

Class The way it works Help restrict What it means for you
Class 1 Pedal-assist motor solely Stops aiding at 20 mph Normally allowed on bike paths and trails
Class 2 Throttle and pedal help Stops aiding at 20 mph Throttle use could also be restricted in some areas
Class 3 Pedal-assist motor solely Stops aiding at 28 mph Generally banned from shared-use paths and should have age and/or helmet guidelines

Hub motor vs. mid-drive: What’s the distinction?

Most folding e-bikes use hub motors, normally constructed into the rear wheel. They’re easy, reasonably priced, and beginner-friendly. The facility comes on easily—extra like a gradual push than a sudden surge—which makes them nice for informal riders or flat metropolis commutes.

Mid-drive motors, which sit on the cranks, are extra environment friendly and provide higher torque for climbing hills or hauling masses. They really feel extra like conventional biking as a result of they work together with your gears. You’ll discover them much less typically on folding bikes resulting from their price and complexity, however when you do, it’s a premium improve.

Motor energy is normally measured in watts, starting from 250W to the authorized U.S. max of 750W. Extra watts can imply extra pace and stronger acceleration, however torque is what actually helps with hill climbs and fast takeoffs. When you’re in a hilly space or need zippier begins, search for motors with excessive torque rankings.

Batteries: Vary, charging, and what actually issues

Vary is a kind of options e-bike makers like to hype—and it’s getting higher on a regular basis. As of late, even folding fashions typically boast 60, 70, and even 85 miles on a single cost while not having a second battery. However truly hitting these numbers? That relies upon.

Revealed ranges normally listing a minimal and a most, and the real-world quantity lives someplace in between. Terrain, rider weight, cargo, temperature, tire strain, and the way arduous you push the motor (pedal help vs. throttle, low vs. max degree) all make a distinction. For instance, my commute in the summertime takes 1 / 4 of my absolutely charged battery, however in excessive headwinds or freezing temps, it may possibly simply eat up greater than half of my battery. Similar trip, similar battery, completely different situations. In case your deliberate trip is 15 miles round-trip, don’t purchase a motorcycle with a 15-mile max and hope for the most effective—get one thing with some buffer.

Detachable batteries are particularly useful for folding bikes. You’ll be able to cost them inside, swap in a spare for lengthy rides, or take away them for storage. Built-in batteries preserve issues smooth and tamper-resistant, however imply bringing the entire bike to an outlet. Larger batteries weigh extra, however additionally they offer you extra flexibility and fewer vary nervousness.

One very last thing: battery security isn’t non-compulsory. Search for certifications like UL 2849 (complete system), UL 2271 (battery solely), or EN 15194 (European customary). These present that the battery’s been examined towards overheating, influence, and common misuse.

FAQs

Q: What are the downsides of folding electrical bikes?

Folding e-bikes will be game-changers for commuters, however they arrive with a number of trade-offs. They’re normally dearer than non-electric bikes and heavier than customary folders. You’ll additionally want a spot to cost—or search for a mannequin with a detachable battery you may cost indoors. Over time, batteries degrade and have to be changed. And whereas most bike outlets can service customary elements, proprietary programs or wiring may have brand-specific assist. Folding frames additionally introduce further put on factors, comparable to hinges, latches, and clamps, which can loosen over time and with repeated use.

Q: Can folding e-bikes go lengthy distances?

Sure. Many folding e-bikes promote as much as 40 to 80 miles of vary, however real-world numbers rely on terrain, help degree, rider and cargo weight, and climate. When you trip largely flat roads on low help, your battery will stretch a lot farther than when you’re utilizing throttle up steep hills. As for consolation, smaller wheels and minimal suspension imply you’ll really feel extra of the highway, so “lengthy distance” may include extra bumps.
 

Q: How a lot do folding electrical bikes weigh?

Folding e-bikes usually weigh between 35 and 70 kilos. Lighter fashions made with carbon fiber or smaller motors can dip under 35, whereas extra highly effective or accessory-loaded fashions hit the higher finish. They’re nonetheless simpler to maneuver than conventional e-bikes because of their compact form and smaller wheels—however lifting one right into a trunk or up stairs? That’s the place the true check is available in.

Q: Are folding e-bikes allowed on trains and buses?

In most cities, sure. Transit programs normally permit folding bikes on board, particularly in the event that they’re compact and folded earlier than boarding. Simply be ready to hold or roll it rapidly, and keep away from peak commute instances if the bike takes up house. At all times examine native transit guidelines—some programs limit e-bikes by weight, battery dimension, or class.

Ultimate ideas on the most effective folding electrical bikes

Folding electrical bikes are a humorous form of transportation. They’re purpose-built to be simple to move or retailer, which can imply some compromises in driving consolation. Nevertheless, this class of bikes has come a great distance, incorporating fatter tires, extra suspension programs, and hydraulic brakes for superior stopping energy, together with extra choices than ever earlier than. Discover the mannequin that matches—actually—into your condo, automobile, or commute, and benefit from the trip. 

 

Outdoor gift guide content widget

2025 PopSci Out of doors Reward Information

 

What’s !vital #3: Popover Context Menus, @scope, New Internet Platform Options, and Extra

0


The developer group hasn’t wasted any time kicking off 2026 with some actually nice articles, demos, and insights. Firefox 147 and Chrome 144 additionally shipped, and whereas they’re not jam-packed with options, the releases are nonetheless fairly thrilling for what’s usually a sluggish time of yr, so with out additional ado, right here’s what’s vital from the final couple of weeks (or ought to I say the first couple of weeks, of 2026?)…

Chris Coyier (a well-known title, maybe) exhibits us learn how to construct context menus utilizing popovers and anchor positioning over at Frontend Masters. Curiosity invokers,

, discrete transitions, @starting-style, and fallback positions are additionally talked about, so seize a pickaxe, as a result of this one’s a little bit of a goldmine.

Additionally, anchor positioning went baseline this week, so you should use it on manufacturing web sites now! Do we’ve got our CSS characteristic of the yr already?

Scoping CSS with @scope

Funnily sufficient, I additionally obtained the chance to write down one thing for Frontend Masters, and I went with @scope. @scope has been my most-anticipated CSS characteristic for fairly some time now, and Firefox delivery it of their ultimate launch of the yr (making it baseline) made it my characteristic of the yr, so I’m very joyful to kick off 2026 with this little how-to on utilizing @scope and scoping CSS total.

Producing gradient borders from a picture supply

On this demo, created and posted by Ana Tudor on Bluesky, Ana blurs a picture and masks it with a border. You possibly can really accomplish this in Safari utilizing simply three strains of CSS, however the cross-browser resolution isn’t too advanced both (the important thing elements are the backdrop-filter and masks CSS properties).

Given the present reputation of gradients, blurs, and dare I say it, glass, it’s a fairly candy impact that you could most likely adapt for different situations.

Offset gradient border from img supply – how would you get the end result from the display screen under? Actual hole transparency, border gradient obtained from the picture.

My options on @codepen.io:

Safari solely in 3 declarations codepen.io/thebabydino/…

Cross-browser
codepen.io/thebabydino/…

#CSS #filter

[image or embed]

— Ana Tudor (@anatudor.bsky.social) 11 January 2026 at 09:52

You most likely don’t want tabs

HTML, like CSS, is soooo good now. That being stated, regardless that we’ve been getting all these new HTML parts that allow us to construct interactive elements with out JavaScript, that doesn’t essentially imply that we must always. Stephen Margheim says that tab elements are over-engineered more often than not, and explains why and what you are able to do as a substitute.

A sizzling take after seeing one more fancy tabs design:

the traditional “tab element” is over-engineered for 90% of use instances.

You most likely do not want it…

— Stephen Margheim (@fractaledmind.bsky.social) 3 January 2026 at 19:57

Utilizing your OS as a CMS

Talking of simplicity, Jim Nielsen launched me to this extremely cool OS-as-a-CMS idea as he explains how he obtained “Edit Publish” buttons on his web site to open the native doc on his laptop in iA Author, fully negating the necessity for a CMS. Jim walks you thru the entire thing, however the important thing ingredient is just a bit hyperlink with a customized URL scheme:

Edit

I really like this as a result of I additionally write in Markdown (utilizing iA Author, no much less), and may will simply combine this into my Eleventy construct. Nevertheless it obtained me considering: do some other apps have their very own URL scheme? Properly, because it seems, a few of them do! Right here’s an incomplete listing (with examples of ‘edit’ instructions for every app):

  • Obsidian: obsidian://open?vault=posts&file=publish
  • VS Code: vscode://precise/path/to/publish.md:9:1 (:9:1 is the road and column quantity)
  • Ulysses: ulysses://x-callback-url/open-item?id=ITEM_ID
  • Elegant Textual content (with subl protocol): subl://precise/path/to/publish.md:9:1
  • Apple Shortcuts: shortcuts://run-shortcut?title=Edit&enter=/path/to/publish.md (nice for doing stuff with apps that don’t have customized URL schemes)

Fast hits and new internet platform options

As (hopefully?), we publish Fast Hits all through the week. One of the simplest ways to search out them is within the sidebar of the homepage, and so they’re both hyperlinks to issues that you could learn in only a minute or two, or simply PSAs to learn and bear in mind. Anyway, right here’s what you may’ve missed:

Prepared for the primary cool demo of the yr?

A mini Mario world with keyboard management. Sure, you may transfer Mario and in addition leap! 👀

Demo: codepen.io/t_afif/full/… through @codepen.io

✔️ 100% CSS Magic
✔️ Minimal HTML
❌ 0% JavaScript

A Chrome-only experimentation utilizing fashionable CSS.

[image or embed]

— CSS by T. Afif (@css-only.dev) 2 January 2026 at 13:39

And eventually, listed here are my high picks from what Firefox and Chrome shipped on Tuesday:

Thanks for tuning in. I’ll see you in two weeks! Be there or be sq. (aspect-ratio: 1)!