ZXDB Guides

[PeterJ]

@Einar Saukas,

If we add files to an entry which is in ZXDB, and is marked as MIA do we need to make a 'small change to entry' to change it to available, or is this done automatically?

Example here:

viewtopic.php?f=30&t=7745

https://spectrumcomputing.co.uk/entry/1 ... x_issue_07

Thanks

PS - I appreciate you went through all the existing titles following an earlier request. Thanks for doing that!

[Einar Saukas]

If we add files to an entry which is in ZXDB, and is marked as MIA do we need to make a 'small change to entry' to change it to available, or is this done automatically?

It's automatic now.

PS - I appreciate you went through all the existing titles following an earlier request. Thanks for doing that!

You are welcome!

[Einar Saukas]

About the Spectrek book:

I got it! (Not at my parents)



And here is the year and all:



So what you reckon?

It's the same year as the magazine type-in so we cannot be sure which one is the original publication.

Was the book mentioned in the magazine articles or vice-versa?

[Arcachnar]

In a previous post I said that I could have sworn that I saw an article about the book Spectrek....and well.....you know...
|
|
|
|
|
|
|
V
SURPRISE!!!

I have found the article that I was looking for:

https://worldofspectrum.org/archive/mag ... er/11/2/30

This means that the articles and the type-in the Dutch magazine "Sinclair Gebruiker" came indeed first and then the book.

Translated: The first three issues (from November 1984 to November 1984 to January 1985) of "Sinclair User" featured a series of articles by that name. Author Walther Schoonenberg described the principles of structured programming on the basis of the construction of an exciting strategy game, written in BASIC. Game and text have now also been published in book form in an expanded form.

Book: 1985 for ƒ19,75 (fl.) or Gulden or Dutch Guilder(s).

If I want to give the price of the book I should do this with "Small changes for existing ZXDB title/entry", right?

Edit: And how is it done when I want to insert different prices for a game/book that is sold/published/made in different countries like this;

The Antagonists £5.95 (Book)

De Antagonisten ƒ 30,- (Book) ƒ 19,50 (Cassette)

The Secret of Arendarvon Castle £5.95 (Book) £5.70 (Cassette)

Het Mysterie van Arendarvon Castle ƒ 30,- (Book) ƒ 19,50 (Cassette)

Also:

[Einar Saukas]

Thank you!

I just added this new information. Please wait for the next ZXDB update, then take a look.

[PeterJ]

Hi Einar,
I've been given a type in from Computing Today. The magazine is in the archive, but there are not individual pages scanned for the magazine viewer. Do I still add the magazine references as normal?

Thanks

[Einar Saukas]

In this case, you will have to add a new entry without original publisher and release date. Then later use "small changes for magazine references" to add original type-in information.

[pavero]

Well, if we want to see the individual pages at SC, we need the magazine in JPG format (one file per page), not in PDF.

@Einar Saukas
What is the correct structure for these files - correct filenames, all files (issues,pages) in one directory, etc.?

BTW for which magazines do we already have individual pages?

[PeterJ]

@pavero,

The magazine I had a listing for was Computing Today.

[pavero]

@pavero,

The magazine I had a listing for was Computing Today.

It seems the magazine already has the structure for individual pages, only the links are broken?

[Einar Saukas]

What is the correct structure for these files - correct filenames, all files (issues,pages) in one directory, etc.?

Instead of storing an individual link to each page of each magazine issue, ZXDB simply stores a generic mask per magazine. For instance this is the link for Computing Today pages:

/pub/sinclair/magazines/ComputingToday/Issue{y2}{m2}/Pages/ComputingToday{y2}{m2}{p5}.jpg

This link points to the available pages stored at Martijn's WoS mirror.

It seems the magazine already has the structure for individual pages, only the links are broken?

It's not broken. See for instance here:

https://spectrumcomputing.co.uk/issue/1 ... y/1983-May

Whenever the page link is red, it simply means Martijn's WoS didn't have this magazine issue.

We don't talk much about the magazine section of ZXDB, but we have been doing a lot of work there. ZXDB currently contains almost 10,000 paper magazine issues. About 7,400 of them are available online. That's more than twice of Martijn's WoS which had about 3,400 paper magazine issues online. Therefore most links to Martijn's WoS pages will be red.

[Einar Saukas]

Well, if we want to see the individual pages at SC, we need the magazine in JPG format (one file per page), not in PDF.

Yes. However I have a few concerns:

• We have about 7,400 available magazine issues. Assuming 100 pages per issue on average, it means about 740,000 pages. And these numbers are always increasing. Do we have a volunteer to organize roughly a million pages?

• Assuming 200Kb per page scan on average, it means almost 150 Gb. Where are we going to store it?

• Is it really worth it? We already have PDFs for these magazine issues. Users just need to click on a link and the PDF will open directly in the correct page. Afterwards they can easily browse pages, download the entire PDF to view offline, etc. I don't think all this extra work in adding individual pages would add much benefit to them. I would rather invest our time improving other areas instead.

[PeterJ]

@Einar Saukas, @pavero,

Archive.org seems to create image files of each page in a PDF. This is an issue of Computing Today.

https://ia902504.us.archive.org/view_ar ... 12_jp2.zip

Click on the 'JPG' link for the page you want to view.

Could our internal page viewer not link to these?

Personally I prefer viewing magazines in our page viewer, but maybe that's just me!

[pavero]

• Is it really worth it? We already have PDFs for these magazine issues. Users just need to click on a link and the PDF will open directly in the correct page. Afterwards they can easily browse pages, download the entire PDF to view offline, etc. I don't think all this extra work in adding individual pages would add much benefit to them. I would rather invest our time improving other areas instead.

Good point! The main reason should be to reduce loading times of these files. There is much quicker to display 200KB jpeg than 10MB PDF. But how I can see now, the online PDF viewer at Archive.org works quite fine, it shows the current page firstly before the others. So it seems probably useless now to have two different models for magazines (both JPG and PDF versions).

But the one thing still left ... the internal page viewer is functional for limited number of magazines now ...

[pavero]

@Einar Saukas, @pavero,

Archive.org seems to create image files of each page in a PDF. This is an issue of Computing Today.

https://ia902504.us.archive.org/view_ar ... 12_jp2.zip

Click on the 'JPG' link for the page you want to view.

Could our internal page viewer not link to these?

Personally I prefer viewing magazines in our page viewer, but maybe that's just me!

Interesting! It would be good to investigate this in more detail!

[PeterJ]

I don't know the technical reasons for this, but when using mobile devices (Android) I find our viewer provides a far crisper image when 'pinching to zoom' compared to when using the archive.org viewer

[PeterJ]

The image links from archive.org seem to work rather well.

[pavero]

Here is the test page:
https://spectrumcomputing.co.uk/page2.p ... 289&page=1

It works fine for me. I think every magazine with PDF version at Archive.org can have their individual pages in JPEG too.

Final implementation I would pass to Einar who knows all logic behind the PHP scripts for magazines.

There is only one problem. Some magazines/issues have really bad resolution in their JPEG format at Archive.org. But we can't do anything with it.

e.g.
https://spectrumcomputing.co.uk/page2.p ... 332&page=8

[PeterJ]

Excellent @pavero,

Thank you.

[StooB]

There is only one problem. Some magazines/issues have really bad resolution in their JPEG format at Archive.org. But we can't do anything with it.

e.g.
https://spectrumcomputing.co.uk/page2.p ... 332&page=8

The solution to this would be a script that downloads the jp2 file to a temporary location and converts it into a jpg in the background using ffmpeg.

This will produce higher resolution jpegs than the ones stored at archive.org, but it would increase server load, and you'd need to regularly delete all the files to avoid filling up the server.

[PeterJ]

Here is the test page:
https://spectrumcomputing.co.uk/page2.p ... 289&page=1

Final implementation I would pass to Einar who knows all logic behind the PHP scripts for magazines.

Hi @Einar Saukas,

Did you have any thoughts on this?