RSS Feeds

Scientists and sky gazers will watch intently as rocket’s upper stage slams into lunar surface on Wednesday

A drifting SpaceX rocket is on a collision course with the moon after launching a pair of lunar landers more than a year ago.

The rocket’s upper stage will unintentionally slam into the moon on Wednesday, carving out a crater and sending up a plume of dust and rubble that scientists – and sky gazers – are eager to observe.

Контекст: я пилю сервис туннелей (грубо — свой ngrok: берёт локальный порт и вешает на него публичный адрес с валидным TLS). В какой-то момент дошёл до денег — прикрутил онлайн-кассу, чтобы люди могли оплачивать платные тарифы. И тут же упёрся в классическую стену: кассу невозможно протестировать на localhost.

Проблема в том, что касса общается с тобой асинхронно и по своей инициативе. Пользователь оплатил на стороне платёжки — платёжка шлёт тебе вебхук (IPN, callback, notification — называют по-разному): «заказ такой-то оплачен, выдавай услугу». Этот запрос идёт снаружи к тебе. А у твоего ноутбука нет публичного адреса, и касса при всём желании не достучится до http://127.0.0.1:8080/api/tariff/crypto-callback. Пока callback-URL недоступен из интернета, ты протестируешь ровно ничего дальше кнопки «оплатить».

Обычно это лечат деплоем: залил на стейдж, потыкал, нашёл опечатку в парсинге тела вебхука, снова залил. Цикл в несколько минут на каждую однострочную правку. И вот я сижу, смотрю на этот цикл, смотрю на свой недоделанный сервис туннелей, назначение которого — прокидывать локальные порты в интернет ровно для таких случаев… и до меня доходит, что инструмент для решения моей проблемы я держу в руках. Точнее, пишу.

Exclusive: Danielle Kirton calls on Ed Miliband to help her get answers about why evacuation order came too late

The bereaved daughter of a British couple who died in a Spanish wildfire is calling on the foreign secretary, Ed Miliband, to get answers from Spain’s government about why her parents were not told to evacuate until it was too late.

Severe wildfires have burned across Spain and France since early July, killing people, destroying thousands of homes and forcing more than 300,000 people to flee.

Celebrities including Charli xcx and Gabbriette and avant-garde brands honour tradition while also rejecting it

If a wedding day brings to mind a bride dressed in a big white dress, the styles coming down the aisle in the near future may strike you as a little different. Whether it is through celebrity references, or new brands, wedding gowns are entering a more experimental era.

For Charli xcx’s marriage to the 1975 drummer George Daniel last year, the bride wore a minidress and sunglasses.

Привет! Меня зовут Вадим Вахрушев, я старший DS-инженер в Авито, занимаюсь задачами поиска. В статье расскажу, как устроено ранжирование в Авито Работе: какие ML-модели помогают создавать выдачу вакансий, и какие инсайты мы вынесли из нескольких показательных АБ-тестов.

Думаю, статья будет в первую очередь интересна ML-инженерам, которым важно разобраться, как поисковое ранжирование адаптируется под конкретную бизнес-вертикаль. Если хочется узнать про поиск в Авито в целом, загляните в обзорную статью — «Как работает поисковое ранжирование для миллионов объявлений Авито». В этом материале я сфокусируюсь именно на специфике вакансий.

Сейчас в группы тг стали выкладывать вакансии со ссылками на Linkedin, где я перманентно заблокирована, разблокировать не могу из-за того, что российское гражданство, а все новые аккаунты они видимо метчат по опыту работы и тоже отправляют в бан.

И вот появилась вакансия “Middle ML-Product Manager” в Купер. ML я люблю, и занималась ценообразованием в Х5, так что сферу знаю. Только ссылки все в LinkedIn - откликнуться не могу.

Но я упорная - пошла искать.

На сайте Купера вакансии нет, как и вообще вакансий продактов.

Пошла искать через гугл. Нашлись несколько закрытых вакансий на сайте Купера и на hh. 

На hh нет ни одной вакансии продакта в Купере.

Зато нашлась вакансия по следующим ссылкам:

Nebraska's Medicaid director, in an exclusive interview with nonprofit news organization Tradeoffs, said roughly 200 people will lose health coverage on Aug. 1.

(Image credit: Rebecca S. Gratz for The Washington Post)

Трафик из поиска падает, а доход паблишера все еще можно увеличить. Если разобраться, как устроен аукцион, который продает рекламу на сайте. Разберу это простыми словами, а также почему одна рекламная сеть — это потолок, и как на сайте за 4 дня удалось увеличить доход на 159%, ничего не меняя в трафике.

Всем привет! Решил попробовать новый формат на хабре: делиться результатами своей работы за прошедший месяц. С интересными ссылками на события, проекты, обсуждения, релизы из мира питона, моего телеграм канала и нашего замечательного чата (куда много людей приходят пиарить свои проекты).

Давайте сразу договоримся: я не публикую ИИ слоп, пишу статью ручками (потому что уважаю своего читателя) и рассказываю интересное и актуальное из своей бесплатной работы в опенсорсе, а с вас лайк статье (если она будет для вас интересной).

Погнали смотреть, что у нас в питоне происходило!

UK prime minister is the first head of a major nation to call on the under-pressure Fifa president to step down

• UK politics live – latest updates

Andy Burnham has called on the Fifa president, Gianni Infantino, to resign over his “outrageous” plans to sell a stake in the World Cup.

Infantino faces growing pressure over his proposals, which triggered a backlash among football bodies and prompted the resignation on Friday of a key adviser.

Мы перешли в опытно-промышленную эксплуатацию. Контуры работали, пользователи были подключены, подразделения собирали потребность не в Excel, а в единой системе. На управляющем комитете я именно об этом и рассказывал.

Спонсор выслушал и спросил: а что скажут закупщики? Им легче стало работать?

У меня не было хорошего ответа.

AI code completion built human review into the process by design. A developer types, a suggestion appears, and a human decides whether to accept it. A person looked at every line before it shipped.

Agentic AI breaks that review loop. An agent can open a merge request, call a tool, modify a CI/CD configuration, and push a change, sometimes without a person reviewing each individual step. Add in the Model Context Protocol (MCP), which lets agents connect to external tools and data sources on their own, and the question engineering leaders are asking shifts. It is no longer "which model writes the best code?" It is "what is this agent allowed to do, and how can we prove what it did?"

GitLab's own research across more than 1,500 developers and technology leaders found that 73% of respondents are concerned about the long-term maintainability of code and 86% agree that without clear governance, AI-generated code can compound technical debt faster than traditional development practice.

In this article, you'll learn how to proactively address some of the challenges organizations are starting to see with governing agentic AI. We'll introduce you to a governance framework for agentic AI in software development and explore what to control, where human review still belongs, how to measure a rollout, and a practical checklist for teams standardizing on GitLab Duo Agent Platform.

Why agentic AI needs a different approach to governance

The security model for an interactive coding assistant is straightforward because a human is in the loop at every step: a developer asks a question, reviews a suggestion, and accepts or rejects it.

Agentic AI in automated workflows requires a different approach to governance. When an agent can run tests, modify configurations, and take multi-step actions across the software delivery lifecycle without a human reviewing each step, the relevant questions change to:

• What can this agent access?
• What is it authorized to do?
• What actions did it take, and can that be proven after the fact?

Most teams are already feeling this challenge. Ninety-two percent of DevSecOps professionals report some governance challenge with AI-generated code, and the specific concerns point directly to the questions above.

The top concerns include:

• Code attribution: The ability to tell which code was AI-generated versus human-written in the first place.
• Traceability to intent: Connect AI-generated code back to the business requirement it was meant to satisfy.
• Documentation that scales: Manual documentation practices don't hold up once agents are generating a growing share of the codebase.

Once an agent is approved for a project, it can typically write, delete, and push changes without someone reviewing the action before it happens. However, you are still accountable for what lands in the codebase regardless of whether a person or an agent made the change.

Governance for agentic AI is not an add-on to code completion governance. It is a different approach, built around identity, permissions, and auditability.

Setting controls for MCPs, agents, model access, and tool permissions

Once agents can call tools and connect to external systems through protocols like MCP, permissioning becomes the control point. A useful governance model answers three questions before any agent runs:

1. Which agents and flows are allowed?
2. Where are they allowed to operate?
3. Which models can the agents use?

In practice, this looks like a few layers working together:

• A central catalog for agents and flows: Rather than every team standing up its own agent integrations, a shared AI catalog lets administrators decide what gets published, aligned to the organization's existing roles and group structure.
• Composite identity: Every AI agent's identity should be linked to the human user who requested the action, so activity is never attributable to the agent alone. When an agent tries to access a resource, both principals, the agent and the human who instructed it, need to be authenticated and authorized before access is granted.
• Tool approval guardrails: Individual agent tools can be set to run autonomously, pause for a human reviewer, or stay blocked outright, so a sensitive action like writing a file or deleting a resource waits for sign-off before it executes.
• Prompt guardrails: Because agents can process untrusted input (a webpage, an issue comment, a file an attacker controls), the platform needs to detect attempts to hijack agent behavior mid-workflow, not just log it after the fact.

The goal is a control plane that treats agent permissions the same way an organization already treats human permissions: role-based, auditable, and consistent across every project.

Data privacy and self-hosted AI: The questions worth asking

Source code is one of the most sensitive assets an enterprise has, and every AI feature that reads it raises a data-handling question. Before rolling out agentic AI broadly, engineering and security leaders typically want clear answers to a short list of questions:

• Does the vendor train models on our code?
• Who owns the inputs and outputs?
• Where do our subprocessors sit, and can we be notified when that list changes?

For organizations in regulated industries, the answer to "where do our subprocessors sit" often needs to be "nowhere outside our own infrastructure." That is why self-hosting matters as a governance lever, not just a deployment preference. Self-hosted options let a team run its AI agents entirely on infrastructure it controls, while still accurately tracking team usage and satisfying regulators.

Bring-your-own-model support extends that further, letting administrators connect models they have already validated internally and map specific models to certain agent flows. With this method, a sensitive workflow can be pinned to a model the organization trusts while less sensitive workflows use a managed option.

Human-in-the-loop: Decide where review still belongs

Governance does not mean blocking agents from acting autonomously. It means deciding, deliberately, where autonomy ends and review begins. A workable policy usually separates two modes:

• Interactive work, where a developer is present, sees each suggestion, and approves or rejects it directly. This is the mode most teams already understand from AI code completion.
• Automated or headless work, such as agents running inside CI/CD pipelines without a developer watching in real time. Here, the review has to happen either before the action (through tool approval guardrails) or immediately after (through an audit trail a human can inspect).

Key decision points worth defining

Code review, testing and validation, and deployment approval should all be defined by answering questions like:

• Which of these can an agent complete unassisted?
• Which require a named human sign-off before the change proceeds?

In practice, each checkpoint needs its own enforcement mechanism. This could look like:

• Merge request approval policies determine who has to sign off before a merge request lands, regardless of whether an agent or a developer opened it.
• Tool approval guardrails for agents decide, tool by tool, whether an agent's action runs autonomously, pauses for review, or stays blocked.
• Scanner enforcement holds a change at the pipeline level until it clears the security and quality checks your organization requires.

We recommend creating an organization-wide AI governance policy rather than leaving it to individual team norms. This helps create consistent usage and policies, and helps auditors verify your use of AI.

Added context from the decision layer

There is a longer-term benefit to capturing these approvals as structured records rather than letting them evaporate in a chat thread or a reviewer's memory. An exception granted last quarter, the policy version it was granted under, and who approved it are exactly the kind of organizational judgment that both auditors and future agents need to reference.

Enterprises that treat decisions as durable, queryable events get the added benefit that the next reviewer, human or agent, starts with additional context instead of starting from zero.

Learn more about the decision layer and how capturing decision events can impact your software development.

5 metrics that measure an AI rollout

Five categories are worth tracking from the start, and they should be reviewed together rather than in isolation, since a spike in adoption without a matching look at risk metrics is itself a warning sign.

1. Adoption: Active users of AI features week over week, number of flows or agents run, and which teams have turned agentic capabilities on versus which have not.
2. Acceptance and quality: Downstream signals like revert rate on AI-assisted changes and how often AI-authored merge requests pass review without rework. A single acceptance-rate number is a weak proxy for effectiveness on its own, since it says nothing about what happens to the code after it is accepted.
3. Risk: Track how often tool approval guardrails pause an action for review, how often those pauses result in a blocked or modified action, and whether any agent activity triggers a policy violation.
4. Remediation: Keep an eye on the scanner coverage across projects, the share of vulnerabilities that get auto-remediated versus manually triaged, and time to resolve findings once flagged.
5. ROI: Once you have a full quarter of data, explore the ROI of your AI investment. Track how much time developers save on tasks agents now handle and weigh that against the cost per resolved issue or merged change, in credits or compute spent. Check out this tutorial for an in-depth look at your AI ROI — you’ll learn how to transform raw usage data into actionable business insights and ROI calculations.

By tracking all five metrics together, you can catch issues with the AI rollout early. A rollout can look successful on adoption and acceptance while accumulating risk that only shows up in an audit six months later. With the above metrics, you get a holistic view of the value of AI to your organization.

Your GitLab Duo Agent Platform governance checklist

For teams standardizing on GitLab Duo Agent Platform, here is a practical starting checklist before expanding agentic AI beyond a pilot:

• Review the AI Transparency Center and confirm your understanding of data usage, model vendors, and subprocessor commitments.
• Decide, at the platform level, which agents and flows are approved for use, and publish them through GitLab’s AI Catalog rather than letting teams configure their own.
• Set tool approval guardrails as always allow, always ask, or always deny, based on the sensitivity of the tool.
• Set up composite identity so every agent action is linked to the human who requested it, and access requires both to be authorized.
• For regulated workloads, evaluate self-hosted deployment and bring-your-own-model options against your data residency requirements.
• Document explicit human-in-the-loop checkpoints for code review, testing and validation, and deployment approval.
• Turn on audit event streaming for agent activity so every action lands in the same audit trail your organization already reviews.
• Define your rollout metrics (adoption, acceptance, risk, remediation, ROI) before expanding past a pilot, and review them together on a recurring cadence.

Revisit the checklist each release cycle. Governance for agentic AI is not a one-time setup. New agents, new tools, and new models often reopen governance questions.

Where agentic AI speed meets enterprise control

Agentic AI changes what needs governing. Code completion asked a developer whether a suggestion was good. Agentic AI asks what an agent is allowed to touch, who approved it, and whether that can be proven later.

Answering those questions fully requires governance built into the platform itself, not layered on afterward. GitLab Duo Agent Platform provides an AI Catalog, approval guardrails, and audit event streaming directly into the platform where the work happens, so governance isn't bolted on after the fact. Teams get AI-assisted speed and enterprise control together, because the guardrails are part of the workflow, not a separate process running alongside it.

Start a free trial of GitLab Duo Agent Platform. On the Free tier, you can sign up in a few simple steps. If you're already on GitLab Premium or Ultimate, you can turn on Duo Agent Platform and use the GitLab Credits included with your subscription.

29% компаний уже урезали ИТ-бюджеты — и это при том, что сам ИТ-сектор растет и занимает уже 2,2% ВВП России. Вот такая новая реальность: задач становится больше, а рук меньше. 

И тут возникает замкнутый круг, знакомый многим, кто работал в поддержке. Команды сокращают — поток обращений растет. Оставшиеся инженеры тонут в тикетах, выгорают и увольняются. Выход из положения — снять часть возросшей нагрузки. 

В статье разберем шесть инструментов сервис деска, которые помогут делать больше меньшей командой. 

Passports with the president’s portrait, previously only available in DC, will be offered at 27 agencies nationwide

• US politics live – latest updates

Americans outside the nation’s capital will soon be able to get limited-edition passports that feature a picture of Donald Trump and celebrate the 250th anniversary of the founding of the United States, the state department announced.

The department’s 27 passport agencies nationwide will start offering the special travel documents next month. They previously had only been available at the Washington DC passport office for in-person applicants.

Prime minister says he has concerns about one of most significant proposals from Keir Starmer’s government

• UK politics live – latest updates

Andy Burnham is indicating he will scrap plans to curtail jury trials in England and Wales, abandoning one of the most significant overhauls from Keir Starmer’s government.

In perhaps his biggest repudiation yet of the policies he inherited from his predecessor, the prime minister said he had concerns about the proposals and his instinct was to keep juries at the heart of the justice system.

Prime minister says he has concerns about one of most significant proposals from Keir Starmer’s government

• UK politics live – latest updates

Andy Burnham is indicating he will scrap plans to curtail jury trials in England and Wales, abandoning one of the most significant overhauls from Keir Starmer’s government.

In perhaps his biggest repudiation yet of the policies he inherited from his predecessor, the prime minister said he had concerns about the proposals and his instinct was to keep juries at the heart of the justice system.

A Guardian investigation has revealed that a political lobbying firm, founded by the former electoral strategist for David Cameron and Boris Johnson, secretly pays journalists to write pieces for British newspapers that are favourable to its clients.

In a practice that appears to create a back door for private companies to influence public opinion, lobbyists for CT Group, which was co-founded by the campaign strategist Lynton Crosby, told an undercover reporter they had paid journalists to place apparently independent stories in newspapers.

In a statement, CT Group said: 'We do not comment publicly on work we do for clients and we therefore cannot respond to the specific allegations you have made.' CT Group said the events had been 'deliberately mischaracterised' and it did not engage the undercover journalist to do anything and did not pay him anything. It said the journalist was looking for work and 'chased us relentlessly over a period of more than six months'

• Lobbying giant filmed offering reporter payment for flattering client coverage in national press

Вторая часть. В первой я собрал конвейер из ИИ-агентов: он сам ведёт задачу от тикета в YouTrack до коммита - разбирает, правит C++ в проекте на три миллиона строк, собирает и доказывает исправление кадром из живого интерфейса.

Здесь - цифры за девять недель работы на боевых задачах. Например, на код у агентов уходит лишь пятая часть машинного времени. Остальные четыре пятых на то, чтобы доказать, что написанное работает.

Тринадцатикратный рост из заголовка тоже разбираю. Но самое интересное оказалось не в нём.

• Midfielder could still fly to La Manga on Sunday

• Howe says he stood down ‘to recharge and take a break’

Bruno Guimarães was not on Newcastle’s flight to Murcia on Friday lunchtime as Eddie Howe’s former players prepared to meet their new head coach in southern Spain.

As Guimarães, who is believed to be in Brazil, appeared to inch closer towards a transfer to Arsenal, Matthias Jaissle was completing his managerial move from Saudi Arabia’s Al-Ahli. Howe’s successor is expected to arrive at the club’s pre-season training camp at La Manga at some point this weekend.

Passports with the president’s portrait, previously only available in DC, will be offered at 27 agencies nationwide

• US politics live – latest updates

Americans outside the nation’s capital will soon be able to get limited-edition passports that feature a picture of Donald Trump and celebrate the 250th anniversary of the founding of the United States, the state department announced.

The department’s 27 passport agencies nationwide will start offering the special travel documents next month. They previously had only been available at the Washington DC passport office for in-person applicants.

Всем известно, что я предпочитаю использовать для письма перьевую ручку и бумагу, поэтому недавно мне прислали ссылку на вот эту статью и сопутствующее исследование. Я не буду пересказывать их содержание, отмечу лишь самую суть. Когда вы записываете что‑либо от руки, то задействуете большую часть мозга, что определённо идёт на пользу.

Я не особо разбираюсь, как там устроен мозг, но могу сказать, что при письме от руки ему приходится непрерывно решать череду мелких задач. Это и правильное связывание букв, и разделение слов, и расстановка точек над i или чёрточек в t, причём порой по нескольку в одном предложении. К тому же, всё это нужно делать координированным движением не только пальцев, но и всей руки, параллельно обрабатывая абстрактные процессы самого мышления.

Гибридные модели обещают простую вещь. Память под контекст перестаёт расти линейно с его длиной, значит длинный контекст дешевеет, значит на ту же карту влезает больше пользователей. Звучит разумно, но я решил проверить…

Каждую технологическую волну сопровождает один и тот же страх: сейчас всё изменится, старые профессии исчезнут, и все мы останемся без работы. Так было с телеграфом, телефонией, печатным набором, заводским конвейером, потом — с интернетом. И вот теперь с ИИ. Но так ли это?

OWASP, лаборатории, CVE, инструменты и патчи обычно живут в разных местах. Каждый раз собирать этот маршрут заново утомительно, поэтому и сделал RØØT — автономный полигон, который связывает всё в один процесс:

понять → проверить → доказать → исправить → перепроверить

Привет! Я Даниил, DevOps-инженер в KTS.

Недавно я поработал со Stackland. Концепция облачноподобной платформы с кучей коробочных сервисов на своем железе в закрытом контуре давно была любопытна нам (и нашим заказчикам), так что пройти мимо готового решения от Яндекса было бы преступлением. По горячим следам я решил написать эту статью, чтобы поделиться опытом развертывания Stackland на голом железе. Так что ниже вас ждет скорее не обзор, а туториал.

Коротко о том, как это было: сначала я подготовил L2-сеть, бастион с NAT, DNS и NTP, три сервера, образ Stackland и YAML-конфиги. Затем запустил sladm install и через час получил консоль, готовый мониторинг с визуализацией в Grafana и возможность по кнопке в UI поднять managed PostgreSQL, Kafka, ClickHouse и другие сервисы.

Подробнее о том, как это было, рассказываю ниже.

You’ve got to admire the Fifa president’s chutzpah in trying to flog off football’s greatest prize. If only Uefa wasn’t being so pesky about it all

Imagine a plan so distasteful that even the Americans and a Middle Eastern nepo-royal football boss can’t be doing with it. Such is the status of Fifa president Gianni Infantino’s scheme to sell off stakes in the World Cup to private investors. More historically, perhaps, the 55 nations that make up Uefa have reacted to this horror show by voting unanimously to boycott Fifa tournaments, including World Cups, if Infantino presses ahead. That would mean Spain and Portugal boycotting their own men’s World Cup next time around, which you’d think – not to call it too soon – would come off as slightly bad vibes.

A testing moment, then, for the haunted cue-ball. Again. Can it really already be time to turn to Infantino once more, even after he appeared in this spot only on Tuesday, after his 15-page Instagram crash-out about his haters? Yes. Yes, I’m afraid it can. Demanding the federations of the world sell off the World Cup to some fund led by Donald Trump’s son-in-law’s brother will regrettably draw attention to you in ways that will prove difficult to control. As indicated, the Fifa president’s plan has contrived to appear vulgar to the North American confederation, Concacaf, as well as the Asian Football Confederation (AFC).

Marina Hyde is a Guardian columnist

• Research analysis to be delivered by 11 September

• Decision on expansion could be taken this year

Fifa is pushing ahead with plans to expand the World Cup to 64 teams from 2030 and has opened a tender to research companies to assess the economic benefits on a hugely accelerated timescale.

In another twist during a week of astonishing revelations about the global governing body’s plans to reshape football, the Guardian has seen a “research brief” outlining a consultation process to assess the potential increase from the current 48 teams. A period of just four weeks has been allocated for the successful party, which will be selected on 14 August, to analyse the proposed expansion. Fifa has required that the analysis is delivered by 11 September, indicating a decision could be taken this year.

Мне понадобилось разобраться, что происходит на большом проекте внедрения: почему сроки едут и куда девается ресурс. Разработку вела внешняя команда, живьём я её не вёл и контекста не имел — на руках были только следы: трекер, вики, гит, списания времени.

Начал я с самого очевидного — померить, кто из аналитиков сколько делает. Провозился неприлично долго, получил красивые графики и всё выбросил. Дальше — почему выбросил и что нашлось, когда я перестал смотреть на людей.

Сразу оговорюсь про две вещи. Первая: это не статья про оценку эффективности сотрудников, скорее наоборот — про то, почему у меня не получилось её написать.

Вторая — про сам кейс. Он тяжёлый. Спринт вдвое больше собственного лимита, треть незавершёнки без движения месяцами, задачи с семнадцатью сменами исполнителя. Нормальным этот проект не был, и типичным я его не выдаю. Полезен он ровно этим: в здоровой команде все описанные ниже эффекты тоже есть, но дают процентов пять и потому не видны. Здесь они дают триста и становятся различимы. Это стенд, а не образец, и интересен тут метод, а не кейс.

В 19 лет закончил университет, в 22 года основал без опыта свой инвестфонд Situational Awareness, а к 24-м управлял в нем уже $45 млрд с доходностью под 250% годовых. Завидуете этому сыну маминой подруги? Ну, давайте я вам расскажу о том, как и почему он их вчера лишился, при чем тут слово «плечи» – и что отличает ветерана рынка от новичка…

Average pump price reaches 160p a litre, while diesel is up 14.5p to 179p and forecast to increase to 185p soon

• Business live – latest updates

The price of petrol has hit a new Iran war high at UK forecourts, adding to the financial pressure on millions of households about to embark on their annual holiday.

The RAC said that in “very unwelcome news for drivers” the average pump price had reached 160p a litre, reversing the fall to 151p in early July that followed the announcement of a Middle East ceasefire.