DEV Community: SOCAR

What font `2600` mag use ?

SOCAR — Mon, 17 Nov 2025 14:40:34 +0000

Problem

2600 magazine (https://www.2600.com/) been around for few decades, however I've just recently realized that pages and pages of small text (printed edition is about B5 format?) are pleasure to read. Maybe it's because of high quality of materials from interesting people or maybe it's... font. Quick search in generative language models suggest some sort of TimesNewRoman but not exact match. Opening recent issue PDF in Adobe Acrobat Reader lists a hefty amount of fonts that you would need to go one by one and eyeball for a match. So question is what font for given text in PDF is used? There is another way...

Solution

Writing a python script that uses PyMuPDF https://pymupdf.readthedocs.io/en/latest/ quickly gives you which font is used on exactly what piece of text on page of your choice:

import fitz  # PyMuPDF

doc = fitz.open("C:\\c\\2600.pdf")
page = doc[0]  # page 1 (0-based)

text_dict = page.get_text("dict")

for block in text_dict["blocks"]:
    for line in block.get("lines", []):
        for span in line.get("spans", []):
            print(repr(span["text"]), "->", span["font"], span["size"])

Gives you nice view what font is used and where

Answer

2600 magazine uses TimesLTPro size of 8.6pt for their articles. That's it, roll credits.

but...

TimesLTPro-Roman is the PostScript font name of the Roman style in the Times LT Pro family. Vendors and "foundries" (I imagine it's common in some communities to call "foundry" some place font originates) show such font as “Times LT Pro Roman”. Also there's a lot of variations of Times New Roman across operating systems and embedded fonts that are being sold on the market. Lets dive deeper.

PDF File produced by 2600 have quite few fonts embedded. Not sure if all of those are used however I'm focusing only on one we read mostly in the magazine, as previously identified TimesLTPro-Roman

It's object number 619 embedded in the pdf file, however typical attempt on extracting all items from the pdf, does not produce *619 file.

mutool extract 2600.pdf

results in missing id:619 object extraction

To our advantage comes previously used python library: PyMuPDF
After intalling it with pip it also exposes set of oneliner tools that can be used to extract pdf fonts as well

pymupdf extract -fonts 2600.pdf

extracting all 67 fonts (and preserving their names)

CFF

CFF is not natively supported under Microsoft Windows. Compact Font Format and TTF (TrueType) are both outline font formats, but they’re built differently: TTF uses quadratic Bézier curves with TrueType hinting, while CFF is a compact PostScript-based format (often seen inside OpenType-CFF fonts) using cubic Bézier curves. TTF became dominant on Windows and the web, but CFF is still in use because produces smaller file sizes, integrates very well with PostScript/PDF workflows, and has long-standing support in professional publishing and print environments—so a lot of high-end text and type foundry workflows still rely on CFF-based fonts.

To open CFF under windows (or other systems) you can use FontForge https://fontforge.org/en-US/

Thanks to it we can now preview the file and see it's details

Conclusion

In this article I've shown how to establish what fonts are used in PDF file and their copyright owners.

From what it seems, this nice looking font, is a professionally licensed and bought product, from another corporation and used since 2014 at least. From legal perspective it's use anywhere outside of agreed terms would be violation, so anyone considering using it for their own benefit would need probably go one and only "the-bay-way".

Error during VMWare installation

SOCAR — Fri, 05 Apr 2024 14:47:47 +0000

I'm big fan of VMWare Workstation Player, which allows me to test different software configurations and run older operating systems with ease. However sometimes, during upgrade (or reinstall) of the software I hit following error:
Service 'VMware Authorization Service' (VMAuthdService) could not be installed.

Lets leave this error as it is, and fix the problem, so we can later click Retry and continue the install.

It seems, VMware Workstation Player setup guide have some issues when uninstalling previous version of the software, causing mentioned service to remain on the system. Usually it's defunct entry in visible in MMC service snap-in, which we can run by typying services.msc in START -> Run dialog.

Usually that screen will be populated with ERROR2: Cant read description or similar.

To fix this issue and continue with install we have to get rid of the service. To do that we will use sc tool:

Run Command Prompt as Administrator
type in sc delete VMAuthdService and press ENTER key
You should see [SC] DeleteService SUCCESS
Go back to VMware Workstation Player installation and click RETRY

Documentation for sc.exe tool: https://learn.microsoft.com/en-us/windows-server/administration/windows-commands/sc-delete

Quick Guide to Mouse Clicking in C#

SOCAR — Sun, 24 Sep 2023 19:41:17 +0000

Clicking the mouse

Clicking mouse in Windows is done using "Windows API". We will require some methods from Microsoft Windows user32.dll. Also we need to remember that click is a two stage process: first you click the mouse button down, which generate one event, followed by releasing button (second event). Having said that, here's the steps:

Add using System.Runtime.InteropServices; to your *.cs file

Add two methods from user32.dll that we will be using:

[DllImport("user32.dll")]
public static extern bool SetCursorPos(int X, int Y);
[System.Runtime.InteropServices.DllImport("user32.dll")]
public static extern void mouse_event(int dwFlags, int dx, int 
dy, int cButtons, int dwExtraInfo);

mouse_event(?) is method responsible for clicking, while SetCursorPos (?) sets mouse cursor where we want on the screen
We define human readable variables of earlier mentioned two events - LMB_Down and LMB_Up. Its not necessary as we can just provide hex numbers to the mouse_event function, as well.
```
public const int LMBDown = 0x02;
public const int LMBUp = 0x04;
```

Now we make a click at screen position x=128 and y=256

SetCursorPos(128,256);
mouse_event(LMBDown, 128, 256, 0, 0);
mouse_event(LMBUp, 128, 256, 0, 0);

and thats it!
Working example: https://gitlab.com/-/snippets/3602168

Where's my cursor at?

As only few of us have superhuman ability of telling exact x,y position of the cursor, we can use this piece of code (for example running in separate console app) to tell where the cursor is now at. We still have to use "WinAPI" function GetCursorPos (?) which returns position result using out parameter modifier (?) to a struct that we will have to create on our own as well.

Full code of getting the position of the cursor looks like this:

using System;
using System.Runtime.InteropServices;
class Program
{
    [DllImport("user32.dll")]
    public static extern bool GetCursorPos(out CursorPos lpPoint);
    public struct CursorPos { public int X; public int Y; }
    static void Main(string[] args)
    {
        CursorPos position;

        do
        {
            if (GetCursorPos(out position))
                Console.Write($"{position.X},{position.Y}\r");
        } while (!Console.KeyAvailable);
    }
}

gitlab

Data is Null. When "safety" is in the way.

SOCAR — Mon, 05 Jun 2023 17:39:35 +0000

Back in .NET 2.0, which was my starting point of using .NET stack, null was important aspect, around which we were building our software architecture, meticulously crafting code that was NullReferenceException safe.

Nowadays, new safety features of .NET ecosystem introduced Nullable contexts which are here to “save” us from need of additional null checking. Sometimes however they can introduce unexpected behavior causing exceptions in places you would not expect.

The Project

In below example we track and store followers count of profiles in some generic social network.

Visual Studio solution is split up in several projects that reuse code between them. It's outline looks like this:

IFM.BackgroundWorker is gathering data by scraping the website and using IFM.DBA.MSSQL saves it to Microsoft SQL Server. IFM.Shared holds DBO (DataBase Object) used by Entity Framework Context consumed by 3 project:

IFM.DBA.MSSQL - access layer to the database
IFM.BackgroundWorker - uses IFM.DBA.MSSQL directly to save the data
IFM.WebApi - Reads the data using IFM.DBA.MSSQL for downstream clients

The Code

Now lets consider string notes; A string can be null and for decades we used String.IsNullOrEmpty() to check that. The same goes for databases such as Microsoft SQL Server (MSSQL) which allows you to have a NULL value in column of NVARCHAR type

To glue this together in .NET you can use Entity Framework Core (EFCore) with Database First approach creating DbContext and following DBO:

Assuming we have some data in the table, we create a method which filters out reports for specific profile using EFCore.

Finally we write the test, expecting everything to be fine... Just getting data for one of profiles.

And test… fails (as well any implementation of this code in any of the projects). It will throw exception when attempting to create list in GetAllStatsForProfile method and provide relatively cryptic info about this: System.Data.SqlTypes.SqlNullValueException: Data is Null. This method or property cannot be called on Null values.

Reason for this exception is notes column in FollowersData table that allows NULL values. Previously shown FollowersDataDBO class was coded by hand, however using a dotnet ef dbcontext scaffold gives same results – creates class that have notes property not marked as nullable. string type which can be null is not "accepting" null coming from the database

The Problem

At this point we are facing a problem not only because exception is breaking execution of the application but how to address it in regards of architecture. In this example, we control the source code, but what if the project would be closed source?

The Fix

Here's 3 suggestions how to fix the issue:

1. Marking `notes` property of DBO class to accept `null` values

Assuming we can modify the source code of the FollowesDataDBO class we can mark string notes property as accepting null values by adding ? after the type definition.

Interestingly enough the null forgiving operator can't be used here (public string notes {get;set;} = null!) and would not resolve the issue

2. Disabling `<Nullable>`

Disabling null checking by setting in the project file <Nullable> value to disable do the trick as well. However it have be done in project which holds the DBO file (and not projects that reference that project), which affect all other projects that might have logic relying on that setting. In our case we would have to modify IFM.Shared.csproj

Obviously from global perspective we cant predict what impact on projects that reference this and rely on <Nullable> to be enabled

3. Change database schema

Finally, the "worst" approach to fix the issue, that we could introduce, is changing database schema by setting notes column to NOT NULL (Allow Nulls = false) but as projects might be dependent on such schema, the introduced risk of breaking change in this case is big.

Conclusion

In this example I have shown that new safety features might backfire in specific conditions. Proper design of application as well continuous testing is important to catch such errors on early stage of project development. Such safety features might be also a breaking change while migrating from older .NET stack and cause additional workload while mitigating potential risks and architecture changes.

DEV Community: SOCAR

What font `2600` mag use ?

Problem

Solution

Answer

but...

CFF

Conclusion

Error during VMWare installation

Quick Guide to Mouse Clicking in C#

Clicking the mouse

Where's my cursor at?

Data is Null. When "safety" is in the way.

The Project

The Code

The Problem

The Fix

1. Marking notes property of DBO class to accept null values

2. Disabling <Nullable>

3. Change database schema

Conclusion

1. Marking `notes` property of DBO class to accept `null` values

2. Disabling `<Nullable>`