Refactoring is nothing less than an intervention in the code base of systems, which can create stability or generate new risks. The right approach is therefore crucial. IT service provider Avision summarises the most important steps for successful surgery on live code.

In practice, refactoring is much more than a purely technical measure. For improvements to the code base to have a long-term effect, a clear understanding of goals, limitations and processes is required. Based on many years of project experience, Avision has summarised the principles that are important and how refactoring can be integrated into everyday development in a structured manner:

1. Set clear limits and proceed step by step

Refactoring should be deliberately limited and controlled. Instead of large-scale changes, clearly defined, smaller changes that can be verified are crucial. In practice, a step-by-step approach with small commits has proven successful, allowing adjustments to be tested and traced directly. It is equally important to clearly define the scope: Which areas should be optimised and which should not? Without these boundaries, there is a risk that the refactoring project will get out of hand. This not only leads to increasing complexity, but also jeopardises schedules and acceptance within the team.

2. Introduce a comprehensive testing concept

Tests are not a nice-to-have, but a central prerequisite for refactoring to be responsible in the first place. Practical experience has shown that companies benefit from creating tests before the actual refactoring in order to safeguard the existing behaviour of the software. Only this basis makes it possible to see whether changes unintentionally alter functionality. Depending on the criticality of the code area, different types of tests can be considered, with unit tests, integration tests and end-to-end tests being particularly important.

3. Understanding refactoring as a process

Refactoring is most effective when it is not seen as a one-off large-scale project, but as an integral part of continuous software development. This prevents technical debt from accumulating over years and eventually becoming impossible to reduce without considerable effort. A proven approach is to leave the code a little better than you found it with every change. However, if refactoring has been neglected for a long time, a purely incremental approach is often not enough. In such cases, it makes sense to plan additional phases, such as technical sprints or dedicated projects to reduce technical debt.

4. Use AI only in a targeted manner

Artificial intelligence can effectively support refactoring when used deliberately. AI is particularly helpful when it relieves developers of clearly defined tasks, such as creating unit tests, restructuring methods or reducing code duplication. Less effective is the attempt to have large areas of code or entire applications completely optimised by AI using a refactoring approach. Especially in the case of mature legacy systems, technical evaluation by specialists and developers remains indispensable. AI is therefore not effective as an autopilot, but as a supporting tool. Humans remain responsible for architecture, logic and quality.

‘Refactoring is a necessary investment in the future viability of software,’ says Nadine Riederer, Managing Director at Avision. ‘It’s about systematically cleaning up your own code and modernising it in a meaningful way before real problems arise. Sure, refactoring takes time and usually doesn’t deliver any exciting new features. But it creates the basis for further developing software in a stable manner, avoiding technical debt and reliably implementing new requirements.’

This press release is also available at www.pr-com.de/de/avision.

Pressekontakt

Avision GmbH
Christina Karl
Marketing
Bajuwarenring 14
D-82041 Oberhaching
Phone +49-89-623037-967
christina.karl@avision-it.de 

www.avision-it.de     

PR-COM GmbH
Melissa Gemmrich
Sendlinger-Tor-Platz 6
D-80336 München
Phone +49-89-59997-759
melissa.gemmrich@pr-com.de

www.pr-com.de

Ageing IT systems need help. But because modernisation initially costs time, energy and money, many decision-makers recognise the need to act, but are surprisingly creative when it comes to glossing over the facts. IT service provider Avision reveals that behind the sugarcoating lies a failure to provide digital assistance.

Most companies have long been aware that their legacy software will not meet the requirements of the coming years. Technically, this is rarely a secret, and organisationally, it is usually not a secret either. And yet, surprisingly little is often done about it. Instead of making clear decisions, experience from the and the hope is spread that the problem will somehow resolve itself. This is also dangerous because legacy systems rarely explode spectacularly overnight, but slowly but steadily become more expensive, cumbersome and risky. Avision has collected the five most common spurious arguments from practice that slow down active countermeasures.

‘It still works.’

Yes, for now. At first glance, a system that is running often still appears to be stable. But simply functioning has nothing to do with scalability, security and future viability. If there are more users on the system than it can handle, backups have not been tested for years and new legal requirements can only be implemented with a lot of luck or not at all, then the system is no longer really running – it is somehow keeping its head above water. At the latest when an audit is passed only because no one has looked closely, action must be taken. But that is only possible if the system is still alive at all.

‘There is no business case.’

Modernisation often feels like pure money spending without any visible added value; after all, companies are not selling anything new afterwards. However, it is often overlooked that legacy systems make even the smallest change extremely complex, unnecessarily increase the cost of infrastructure and pose security risks that, in the event of an emergency, can cost many times more than the cost of modernisation. The business case rarely lies in the modernisation itself, but in the costs that would continue to accrue without it.

‘It’s so old, no one would hack it anymore.’

Old systems are inconspicuous and, as a result, are often dismissed as uninteresting to attackers. Added to this are arguments such as firewalls, internal networks or a lack of appeal to attackers. This is fatal, because the reality is different: outdated software often means unpatched vulnerabilities, missing updates and dependencies that no one can keep track of anymore. Each factor is a potential gateway for cybercriminals. Security comes from maintainability and transparency – and that’s exactly what neglected legacy systems lack.

‘No one can do that.’

When crucial expertise is concentrated in one or two people, every change becomes a risk. The absence of a single person can be enough to bring systems, processes or projects to a standstill. This problem is often ‘solved’ by leaving everything as it is. It has always worked well so far. The problem is that the longer nothing happens, the more difficult it becomes to build up knowledge or purchase it externally. At some point, a change in the law or a failure will force action – but then under maximum pressure and stress for everyone involved.

‘Never change a running system.’

This phrase sounds like experience and serenity, but it is usually an expression of fear of costs, complexity and the huge amount of work involved. While long-standing employees have learned to live with the idiosyncrasies of the system, new colleagues fail to do so or quickly move on. At the same time, the competition is pulling ahead. A system may still be running today – but without innovation, it will only be lagging behind tomorrow.

‘When it comes to maintenance and modernisation, software is not much different from areas such as road construction or real estate,’ explains Nadine Riederer, Managing Director of Avision. ‘If urgent measures are always put on the back burner and investments are avoided, at some point nothing will work anymore – and then it will become really expensive. There are many excuses, but they cost a lot of time and ultimately money as well.’

This press release is also available at www.pr-com.de/de/avision.

Press contact

Avision GmbH
Christina Karl
Marketing
Bajuwarenring 14
D-82041 Oberhaching
Phone +49-89-623037-967
christina.karl@avision-it.de 

www.avision-it.de     

PR-COM GmbH
Melissa Gemmrich
Sendlinger-Tor-Platz 6
D-80336 München
Phone +49-89-59997-759
melissa.gemmrich@pr-com.de

www.pr-com.de

With the end of the year just around the corner, it’s time for reflection and wish lists. Even in places where there’s usually little room for such things between deployments, ticket floods, and spontaneous “Can you just quickly…” requests. While the fairy lights are going up outside and the city centers smell of mulled wine and gingerbread, IT teams are hoping above all for stable systems and a few quiet days without alarm messages. So it’s time to take a look at the small and large wishes of those who keep IT operations running.

December is a difficult month in IT: On the one hand, many want to complete their projects, but on the other hand, everyone longs for a break from the hectic pace of IT operations. However, these continue unabated at full speed even during the Christmas season. So what’s on the wish list of developers who are trying to find a little peace and quiet between system crashes and last-minute changes during this often turbulent time of year?
We have compiled the most important wishes:

1. Fewer regulations, more fun coding

Every new guideline sounds like bureaucracy bingo: DORA, NIS2, CRA – at some point, no one knows anymore whether they are actually writing software or implementing regulations. Of course, security needs rules, but sometimes you wish the energy spent reading paragraphs could be converted directly into productivity.

2. Audits with long intervals

Christmas joy in its purest form: no checklists, no flood of meetings, no frantic “Who has the minutes from 2022 again?” Between security requirements and compliance updates, there’s hardly any room to breathe anyway. An audit-free year would be like a vacation.

3. A customer who voluntarily increases the daily rate

The probability of this happening lies somewhere between “the server remains stable after patch day” and “the specialist department has no change requests.” But we can dream: a customer who recognizes good work before the controlling department relativizes it—that would be true Christmas magic.

4. A VPN that ALWAYS works

Whether on the train, in the home office, or at the Christmas market – VPNs seem to have an aversion to the festive spirit. As soon as someone plays “Last Christmas,” the connection breaks down. If Santa could really bring one thing, it would be stable tunnels – without timeouts, without curses, without restarts.

5. A firewall that also blocks Christmas stress

Wouldn’t that be nice? A rule that automatically sends all “Quick query about project status” emails between December 23 and January 2 to snowy nirvana. Maybe even with an integrated meeting blocker. Anyone who manages to write this into the next security update would surely win the internal innovation award—at least from their colleagues.

6. A time machine for deployment

For those moments when just one tiny fix suddenly brings the whole system to a standstill. Anyone who has ever had a less than successful deployment on a Friday would give anything to have a time machine to go back 30 minutes before the button was pressed.

7. Finally, all tickets in the green zone

Every project manager’s dream—and a status that all development teams will still be telling their great-grandchildren about. In reality, tickets are not closed, they evolve: bugs become features, features become change requests, and at some point they are given low priority.

8. AI that finally understands what the customer really wants

Not what’s in the specifications – but what’s actually meant. If artificial intelligence could one day interpret customer wishes before they are rephrased three times and escalated five times, it would be the beginning of a more carefree era.

“As diverse as the wishes on our list are, in the end, everyone in IT wants the same thing: good health, good spirits, and a team you can rely on,” says Nadine Riederer, Managing Director of Avision. “Because without these three things, even the most wonderful IT miracles would only be worth half as much. With this in mind: Merry Christmas and best wishes for success and strong nerves to all developers.”

This press release can also be found at www.pr-com.de/de/avision.

Press contact

Avision GmbH
Christina Karl
Marketing
Bajuwarenring 14
D-82041 Oberhaching
Tel. +49-89-623037-967
christina.karl@avision-it.de 

www.avision-it.de     

PR-COM GmbH
Melissa Gemmrich
Sendlinger-Tor-Platz 6
D-80336 München
Tel. +49-89-59997-759
melissa.gemmrich@pr-com.de

www.pr-com.de

Many IT projects fail not because of technology or budget, but because of poor leadership. Between moderation mania, micromanagement and missed decisions, deadlines and motivation often fall by the wayside. Those who bear responsibility but do not take it on cause stagnation. IT service provider Avision presents five types of project managers that everyone knows from practical experience.

Modern IT projects depend on good leadership, decisiveness, empathy and foresight – all attributes that should be combined in project management. That’s the theory. Back in practice, however, IT teams, who are often not to be envied, find very different ‘strengths’ in project management. Decisions are postponed over and over again, responsibility is generously passed on, and meetings are scheduled ad nauseam.

In addition to the many capable and competent project managers in the world, Avision has identified five very specific types who are guaranteed to crash even the most promising IT project into the next wall.

The moderator – always trying

He never chairs a meeting alone and never makes a decision without consulting at least three colleagues. Everyone is allowed to have their say, and they are welcome to do so several times. And at length. His own comments bring even the smallest progress back to square one, and most employees dread the next ‘quick’ team update. After two hours, everything has been discussed, but nothing has been decided. The moderator likes to confuse leadership with physical presence and the length of the discussion with real progress. His strength lies in organising meetings, not in their effectiveness. Projects with him rarely end with a breakthrough, if they end at all. Perfect for those who don’t want to fail – because they never finish. And there’s another small catch: projects become increasingly expensive and the schedule is more of a well-intentioned guideline.

The speaker – a friend of his words

He confuses communication with constant noise. Every meeting begins with a monologue and ends with a repeat of the same. Anyone who tries to contribute triggers an hour-long digression. The speaker loves his own voice so much that he is reluctant to interrupt it – neither for arguments nor for results. At some point, no one in the team talks about problems anymore because they are afraid of provoking a new lecture. As a result, every topic is thoroughly discussed, every project is moderated to death – and every hour becomes one that you will never get back. True to the motto ‘This meeting could have been an email’, the team already looks with unease at the next appointment in the calendar.

The invisible one – camouflage mode on

He is formally the project manager, but in practice he is a myth. He is absent from meetings, offline in chats, and only present as a name in the project plan. While the team organises itself as best it can, the customer wonders about the remarkably long silence. In the best case scenario, the invisible one reappears shortly before go-live, usually to congratulate the team and sell their work as a joint success. For him, leadership is a concept that should work without him. And sometimes it even works, until someone notices that no one is talking to the customer anymore. The invisible one reliably causes escalations – both internally and externally.

The clueless one – a lot of hot air

Content? Nobody needs that! Instead of tackling problems and leading teams with his experience, he counts tickets, checks complicated tables and spends most of his time in mysterious meetings anyway. He takes technical questions ‘with him’ (wherever that may be) – and never brings them back (for whatever reason). For him, the project status is a number, not a state. His understanding of leadership is based on the idea that everyone must update each other at all times – even standing up in the morning. This way, he always remains superficially informed and can update the project plan. He reliably recognises risks only when the worst-case scenario has occurred and successes when someone else explains them to him. He has exactly zero ears for problems in the team and no interest in whatever his colleagues are actually working on in detail.

The dictator – everything stops at his command

He knows everything (better), especially when something is ‘finished’. Criticism is considered disloyalty, contradiction a weakness. Under his direction, tasks are completed rather than done. Pressure is his preferred management tool, and threats are part of his meeting culture. ‘I can see how you prioritise,’ is one of his favourite phrases – half cynicism, half warning. Teams under his leadership only function until the first ones leave. Which, strangely enough, happens very often. After that, he ensures a ‘breath of fresh air’ because new people are constantly joining. He confuses discipline with fear and does not realise that under his command, not only do projects fail, but people also lose motivation.

‘In the end, it’s not just new methods or powerful tools that determine the success or failure of IT projects, but people,’ says Nadine Riederer, managing director of Avision, with a wink. ‘Good project management keeps the team together, provides guidance and makes decisions when others hesitate.’

This press release can also be found at www.pr-com.de/de/avision.

Press contact

Avision GmbH
Christina Karl
Marketing
Bajuwarenring 14
D-82041 Oberhaching
phone +49-89-623037-967
christina.karl@avision-it.de 

www.avision-it.de     

PR-COM GmbH
Melissa Gemmrich
Sendlinger-Tor-Platz 6
D-80336 München
phone +49-89-59997-759
melissa.gemmrich@pr-com.de

www.pr-com.de

Good code is like a tidy home
Overloaded code with unnecessary functions, a lack of structure, and growing complexity: software development also needs to be tidied up regularly to boost efficiency, reduce costs, and ensure traceability. IT service provider Avision shows how the courage to let go of things and a strategic approach can help.

It was a hit series that thrilled millions of viewers around the world: In “Tidying Up with Marie Kondo,” the eponymous host showed how methodical tidying up not only leads to optimal long-term organization, but also makes everyday life easier and happier. Things that many developers can only dream of when looking at their own code. This raises the question: Can this model also be applied to software development? Perhaps not one-to-one, but valuable guidelines based on the principle of “throw away what you don’t need, sort the rest” can already make life a lot easier. Here’s how it works:

Identify unnecessary ballast

Before tidying up, analysis is required. Old systems often contain unused functions that cause technical debt and unnecessarily bloat the code base. Code metrics, test coverage, and usage analyses can be used to identify what can be removed – pragmatically and based on data.

Structure the remaining code

After decluttering, a DDD (domain-driven design) approach ensures order. This puts the domain at the center of the software architecture. By dividing the code into clearly defined bounded contexts, it is broken down into independent, manageable units. This facilitates maintenance, reduces dependencies, and enables targeted further development in line with the actual business logic.

Implementing refactoring correctly

Once the software has been decluttered and the most important components have been structured, the next step is to further optimize the code. This is where refactoring comes in – improving the code without changing the behavior of the system. The aim is to increase the readability and maintainability of the code while minimizing sources of error.

Plan for continuous improvement and maintenance

Software can only be maintained sustainably if order is regularly checked and maintained. Automated testing and continuous integration (CI) help to implement new developments in a structured manner, identify technical debt at an early stage, and maintain high code quality in the long term.

“Good code is like a tidy apartment:
everything has its place, superfluous items are thrown out, and what remains makes life easier,“ says Nadine Riederer, CEO at Avision. ”What developers need for this is, on the one hand, the will and time to declutter digitally, but on the other hand, the courage to throw away the unnecessary once and for all. The results are worth it, because only tidy software is sustainable software.”

This press release is also available at www.pr-com.de/de/avision.

Press contact

Avision GmbH
Christina Karl
Marketing
Bajuwarenring 14
D-82041 Oberhaching
Phone +49-89-623037-967
christina.karl@avision-it.de 

www.avision-it.de     

PR-COM GmbH
Melissa Gemmrich
Sendlinger-Tor-Platz 6
D-80336 München
Phone +49-89-59997-759
melissa.gemmrich@pr-com.de

www.pr-com.de

Cloud technologies are synonymous with innovation and flexibility. But without continuous maintenance, cloud applications can also become outdated, cause maintenance problems and pose security risks. IT service provider Avision shows how companies can implement the modernisation of their cloud systems as an ongoing process.

For most companies, cloud migration is a milestone in their digitalisation efforts. What many easily overlook is that moving to the cloud alone does not guarantee future viability or security. Cloud applications age differently, but they do age – and those who do not actively maintain their strategy will accumulate new legacy issues. This time, not on their own servers, but in the supposedly advanced cloud environment. The speed of technological developments requires constant attention and the right strategy. What should companies bear in mind to ensure that migrated applications do not soon gather dust?

Technical maintenance determines future viability

Consistent lifecycle management of all cloud components is essential to keep systems stable, secure and compatible. Automation and infrastructure as code increase maintainability and scalability, enabling teams to respond more quickly to changes and avoid technical debt. Observability and monitoring tools also help to keep an eye on performance and security at all times.

Without modernisation, downtime and restrictions are a real threat

CI/CD pipelines not only facilitate regular updates, but also often lead to automatic technical modernisation due to limited vendor support. Those who do not regularly review their cloud systems risk applications no longer functioning or failing completely after migration. Especially in the dynamic cloud world with constantly new services and technologies, neglect has a faster and more serious impact than in traditional IT infrastructures.

Regular assessments are a must

To do this, companies should ask themselves a few key questions: Which components and solutions still make sense? What do we really use and what do we no longer use at all? Also important: Has anything changed on the provider side? Are the licences still appropriate and economically viable?

‘The path to a modern and secure cloud architecture does not end with migration, but is an ongoing process,’ says Nadine Riederer, CEO of Avision. ‘It’s similar to road construction: if you don’t carry out continuous maintenance and repair weak points, you’ll soon be faced with a broken infrastructure. It’s similar with the cloud, except that instead of potholes, you have technical debt, security gaps and performance slumps.’

This press release is also available at www.pr-com.de/de/avision.

Press contact

Avision GmbH
Christina Karl
Marketing
Bajuwarenring 14
D-82041 Oberhaching
Phone +49-89-623037-967
christina.karl@avision-it.de

www.avision-it.de

PR-COM GmbH
Melissa Gemmrich
Sendlinger-Tor-Platz 6
D-80336 Munich
Phone +49-89-59997-759
melissa.gemmrich@pr-com.de

www.pr-com.de

Public, private, hybrid cloud, dedicated servers: when modernising legacy software, companies have many infrastructure options beyond on-premises solutions. IT service provider Avision shows which ones are best suited for what.

Infrastructure platforms play a central role in the modernisation of legacy applications. On the one hand, companies can run their applications more efficiently and better meet new requirements by switching from on-premises to cloud environments or dedicated servers. On the other hand, certain platforms can also provide optimal support for specific modernisation tasks. Avision explains the advantages and disadvantages of each option and shows which ones are suitable for which purposes.

1. Public cloud

Public clouds can be used quickly and enable IT resources to be scaled almost indefinitely and in real time. They also save a lot of investment costs, are globally available and, with their AI, IoT and analytics services, ensure a high speed of innovation. However, they also harbour major risks, such as dependence on the provider, data protection issues due to transatlantic data flows and difficult cost control. They are therefore the best choice for rapid prototype development, projects with highly fluctuating loads and non-critical workloads.

2. Private cloud

With private clouds in their own data centre or at a specialised hosting partner, companies retain complete control over their infrastructure and data. Modern technologies such as Kubernetes or OpenStack also give them many of the typical advantages of the public cloud. However, they have to accept higher operating costs, higher initial costs, slower scaling and longer innovation cycles. Private clouds are therefore particularly suitable for applications with sensitive data, high compliance requirements and long-term stable utilisation.

3. Hybrid cloud

This combines the advantages of public clouds with the advantages of on-premises environments or private clouds. Critical data or applications can remain on-pr emises or be hosted in a private cloud, while non-critical workloads are operated in the public cloud. However, the price to pay for this is a complex architecture and integration, higher administrative costs and increased security and governance requirements. The bottom line is that hybrid clouds are suitable for mixed requirements when existing legacy infrastructure is to continue to be used or modernisation is to take place in stages.

4. Dedicated server

Dedicated servers – i.e. physical machines that a service provider operates exclusively for a company – enable complete control over hardware and software. The costs are easy to plan and performance remains constant, as the resources are not shared with other customers. On the downside, high initial costs, longer deployment times and limited scalability are significant factors. Dedicated servers are primarily a good option for computationally intensive workloads, legacy applications that are not containerised or cloud-enabled, and high-security areas such as finance or healthcare.

‘In practice, the best solution is often a mix of different options,’ explains Nadine Riederer, CEO of Avision. ‘For example, dedicated servers combined with private or hybrid cloud models offer a strategic option for particularly sensitive or non-cloud-enabled applications. Companies can physically isolate individual systems while flexibly outsourcing other workloads. In many cases, a dedicated server can also serve as a controlled link between legacy systems and modern cloud components, for example as a secure API node or for migration tests.’

This press release is also available at www.pr-com.de/de/avision.

Press contact

Avision GmbH
Christina Karl
Marketing
Bajuwarenring 14
D-82041 Oberhaching
Phone +49-89-623037-967
christina.karl@avision-it.de

www.avision-it.de

PR-COM GmbH
Melissa Gemmrich
Sendlinger-Tor-Platz 6
D-80336 Munich
Phone +49-89-59997-759
melissa.gemmrich@pr-com.de

www.pr-com.de

Due to the unpredictability of the Trump administration, many companies want to reduce their dependence on US software providers and cloud platforms. IT service provider Avision explains how they can regain their digital sovereignty.

The vast majority of German companies use software from the US and public clouds from US hyperscalers. This makes them dependent on proprietary software from a foreign economic area and gives them no control over the storage and management of their data. In view of the unpredictability of the Trump administration, many companies are now rethinking their strategies and striving for greater digital sovereignty because they fear sanctions or data access by the US government.

Avision highlights the measures companies can take to regain more digital self-determination:

Invest in internal infrastructure

Through targeted investments in internal infrastructure or private cloud solutions, companies can regain control over their data and applications. This is particularly important for sensitive information that falls under the GDPR or contains intellectual property, and for business-critical applications. For these, companies should once again consider on-premises solutions or hosting solutions from trusted IT service providers.

Consider open source alternatives

Unlike proprietary software, open source-based applications give companies full access to the source code. They can check security-critical aspects themselves and ensure that the software does not contain any backdoors through which data could leak. In addition, they are not tied to individual manufacturers and thus avoid the dreaded lock-in effect. They can freely adapt and further develop their solutions and
operate them independently.

Select portable applications

Companies should ensure that their software applications are portable. The applications should support open standards and be container-ready or at least allow flexible integration so that they can function independently of the underlying infrastructure. This gives companies the option of flexibly migrating an application from one cloud platform to another or even to their own data centre, thereby also preventing lock-in.

Rely on widely used technologies

The software applications that companies use should employ established and widely used technologies that are supported by large communities and ecosystems. If rare programming languages or frameworks are used, companies are dependent on IT service providers who have the necessary specialised knowledge. Their freedom to maintain, adapt and further developing the software is then considerably restricted.

‘In view of geopolitical uncertainties, control over your own data and IT structures is increasingly becoming a decisive competitive factor,’ explains Nadine Riederer, CEO of Avision. ‘However, many companies need to take active countermeasures. Those who invest now in the ability to act independently and confidently at all times will strengthen their business resilience.’

This press release is also available at www.pr-com.de/de/avision.

Press contact

Avision GmbH
Christina Karl
Marketing
Bajuwarenring 14
D-82041 Oberhaching
Phone +49-89-623037-967
christina.karl@avision-it.de

www.avision-it.de

PR-COM GmbH
Melissa Gemmrich
Sendlinger-Tor-Platz 6
D-80336 Munich
Phone +49-89-59997-759
melissa.gemmrich@pr-com.de

www.pr-com.de