Zip took 11 years to arrive... And it is Halfbaked

After many blog posts and issues asking to use zip, (something that was already there), Xojo finally add compression methods after a decade.

But they decided to add the feature as part of the FolderItem istead of creating a new Zip Object. What is the problem? they have File Compression but not In memory Compression as a lot of issues asked for :man_facepalming:t2:

They have had a hidden global class
( _GzipString) that does I’m memory compression compression and decompression of strings that may of us have been using for years…

This lets you create hierarchical folder archives… Yes it could have been done in memory but is is a bit of additional capability (though I have plugins that do that so it does not add any capabilities for me.


1 Like

That long…
In this timespan you can build

  • 11 versions of MacOS
  • a 57 km long tunnel crossing the Alps
  • 5 skyscrapers
  • 2 aircraft carriers

A distrust of a significant number of Xojo developers.


I solved this for myself 9 years ago. They literally could have used my code. And, based on what I’ve seen so far, they ought to have:


They had the code for Gzip for far longer than 9 years (actually about 15)
It was just never exposed to users

As Karen noted the code is there an usable just you have to KNOW about it to call it

1 Like

Not invented at Xojo :upside_down_face:

1 Like

neither was the old Gzip :stuck_out_tongue:
I wrote it before I worked there

1 Like

This isn’t 110% unheard off in fairness, for example, Windows eventually acquired the ability to unzip files and Explorer treats .zip files as if they were folders, and this functionality is exposed to the .NET CLR, BUT … the API does not support zip file passwords / decryption. For that you have to use a 3rd party library and so you might as well use it for everything. Fortunately at least one is open source / no cost and works well.

That particular example probably reflects a lack of interest in keeping up with every last feature of the zip format and providing just the basic functionality at the end-user level as well as avoiding the possibility of liability exposure due to having a potentially hackable attack surface by even offering a legit way to bypass protection for legit users. I can almost imagine the memo from the company lawyers!

Of course for Xojo it goes way beyond that. They could at least just state publicly they aren’t going to support it at all and refer users to MB/Einghur, et. al.


OMG and actively encourage and promote a third party ecosystem ???

I really wish they’d focus on those things only they can fix
Like compiler bugs[]=Bug&label_name[]=Compiler&first_page_size=100

Framework bugs[]=Bug&label_name[]=Framework&first_page_size=100

Instead we get things that few people need (like sets) and are something anyone could write