Lars Fosdal

@Vincent Parrett - I agree.  Simply moving to a cloud hosted VM, running your old services, will be expensive as heck. We have metrics to support that statistic.

As for thousands of warnings ... Clean that mess up.  Warnings and even hints can relate to significant problems in the code.  If you have thousands that are "uninteresting", they still can drown out those that are really vital.

np - I forgot to point out that little detail 

I assume you mean runtime warnings from custom code, and not thousands of compile time warnings?

You can create new build configs in the Project Explorer, and they also show up in the Eurekalog Config.



 

Uncheck Base, Check Release - in the EL settings.

Personally I take the 15 secs, so that I can keep track of my leaks 😄

The challenge is often identifying all the ways that an app can activate stuff while the thread is executing.

You could f.x. have a busy flag (or set of flags) you set when you start the thread, and clear on completion, so that when a user opens a menu you can use the flag(s) to disable menu and buttons items that can't be used while the thread is running.

Let's not dive to deeply into the mudhole that is corporate practices.

All we know, is that

• Something went south
• Cleanup takes time
• Practices can only improve
• It is outside our control
• We are short on patience

It's probably smart to inform about the whereabouts and remote working possibilities.

Personally, I would not submit a resume to an "anonymous" position.

1 hour ago, Dalija Prasnikar said:

I wonder who has such landline these days. Our telecom switched to VoIP long ago. So I have a land network connection, but we no longer have classic telephone line.

For phones, there are VoIP solutions, but not really common, so most of us use 4G/5G mobile phones. Even our company phones are mobile based.

Fiber or Cable for land network connections, with 5G mobile internet on the rise.

On 1/28/2024 at 10:02 PM, JohnLM said:

Then we would go back to the days of BBS on dialup 🙂

I actually have a 300 baud modem in my vast archives of useless stuff.
I doubt I would attempt to download 7.5GB over that kind of dialup 😄 but BBSs worked, sort of.
Besides, I no longer have a landline.

Well, my "you" was the general "anyone". And the argument stands: why would anyone chose not to use the massive tooling that already exists for Blazor? Embt would end up in another trying to catch up situation like they did with .net.

It is possible to also capture handled exceptions, but it is costly. 

https://www.eurekalog.com/help/eurekalog/index.php?topic_type_eevents_televraisemeth.php

Who can't resist the call for speculation... 

Why not stay with C# if you want to do Blazor?

That is where the tooling is, and the language is not that different from Delphi. 

If you are using the web-installer, it is probably related to the EMBT Outage.

Are you able to reproduce the effect with a bare bones test application?

Can it be related to DoubleBuffering settings?

Due to the supply chain disturbances, we had a 5 month delivery wait for servers and parts for a while, so I can relate to delays.

The problems here are:

- Lack of proper communication. "No news. Next update at ..." is better than silence!

- Insufficient contingency planning? 

- Insufficient backup and recovery plans?

- No plan B? There are many capable cloud hosting services that doesn't take forever to spin up a pack of servers.

 The cost might be high, sure - but so is the cost of losing the trust of your customers.

AFAIK, there is no automatic deletion of these two services, unless someone intentionally delete their account.

Just to exemplify - I made a gist of your two versions of code. 

https://gist.github.com/LarsFosdal/76bf712c46b3b17d185984d5c0c74494/revisions

Just say the word and I'll remove the gist again.

External code = good.

External binaries = not so much...

I am not saying it is a must, but it makes conversations easier to follow if they have dialog instead of many screenfuls of code.

I am really fond of gists, since you can see the changes, fork them, etc. - and you don't need to log into GitHub to be able to see them.

Things like GitHub gists or PasteBin entries, are great ways of sharing walls of code.

6 minutes ago, ginnix said:

Why is it taking so long to get this working?

That is indeed a good question.

Well, it implied that Daniel did something wrong.  I fail to see the humor.

1 minute ago, Kas Ob. said:

I didn't know that you were oppressed that much by Daniel !

What is that supposed to mean?

DocWiki is still down. GetIt is still down.

Those are pretty crucial.