Downloading a large file from noirlab.edu, only 4.5 GB had been downloaded before "server error" was reported. When
I tried to resume the download, the server would not respond. (and the
issue was not that the server was not reachable - if I started a fresh
download in a different location, it would start - only the resume was
not working.)
Unfortunately, with the noirlab link, we're unable to resume partial
downloads with aria2c after 12 hours or so. The download also seems to
stop after a few hours of downloading, with a server error message.
I've copy-pasted relevant parts of a log file produced by aria2,
showing that the server times out when the client requests either
(a) more than one chunk, or
(b) chunk near the end of the file
2024-02-15 22:09:56.738161 [DEBUG] [SocketCore.cc:993] Securely connected to noirlab.edu (54.200.108.162:443) with TLSv1.2
Reported this to the people at the website, and they made an ip-address-only link available for me to try. That worked, with the resume functionality working even after many hours.
The procedure in this Malayalam video is for changing the state to Kerala, but I suppose a similar procedure would work for changing to another state which is available on the Parivahan Sarathi website.
Notes: Can do renewal and address change in one step.
Parivahan website
Online services -> driving license related
Choose destination state
driving license -> Service on DL (Renewal/Duplicate...)
must be 15 characters when entering DL number - add zeros after year if not.
Nowadays we have grown used to Google Maps having high resolution satellite images. But they are copyrighted images and the Google credits need to be shown onscreen if we use it in a planetarium show or something like that. Looking at OpenSpace and its data sources, googled for some free sources of hi-res satellite images -
I tried to incorporate frame-serving via avisynth+ to my OCVWarp workflow described here, so that steps 1 and 2 could be conflated and the rendering time could be halved - most of the rendering time is taken up by the CPU usage of the codec, I believe. But it did not work.
The latest avisynth+ files work with the latest Virtualdub - even 4096x4096 videos are rendered with no problems. But creating a frameserver with Virtualdub and pointing OCVWarp to the frameserver causes OCVWarp to crash/not respond. AVFS also did not help.
If only an opacity mask (like making the sky less bright) needs to be done, we just need to add the layer on top, and choose the Multiply blend under Compositing.
One of our domains runs their emails on Microsoft365/Outlook, and they were unable to send emails to some other domains they own. Some messages would bounce saying "no such user". It turned out that the recipient domain had been added to Microsoft365 as an additional domain, but no users were added under that domain. Also, that domain was actively being used, with MX records pointing to Google Workspace. So, Outlook or Microsoft365 was not looking at the MX records, but instead looking at the internal users with active directory, and deciding that there was "no such user".
Any other users who had been added to the Microsoft active directory would need to check their inbox on Outlook and not on Google Workspace for any emails sent from our domain on Outlook. This would be relevant for emails being sent from Onedrive, Microsoft Teams, etc.
One of our domain administrators got an email from Microsoft, "Ensure your email authentication records are set up to avoid mail flow issues to third-party email accounts." Apparently, DKIM was not set up properly.
|Microsoft.Exchange.Management.Tasks.ValidationException|CNAME record does not exist for this config. Please publish the following two CNAME records first. Domain Name : ourdomain.org Host Name : selector1._domainkey Points to address or value: selector1-ourdomain-org._domainkey.ourdomain.onmicrosoft.com (This is tested and working fine with dig.)
Host Name : selector2._domainkey Points to address or value: selector2-ourdomain-org._domainkey.ourdomain.onmicrosoft.com (This record returns SOA. I thought there was something wrong with onmicrosoft.com DNS settings. But probably this record is used only when the key needs to be rotated.)
After 90 minutes or so, I checked and found DKIM successfully enabled at that page. The page said that it would take several minutes to roll out the changes, so that DKIM signing would commence after a few minutes/hours.
we were trying to upload special fonts for headings and subheadings in the home page and we could not do it.
Can you pls check if it is related to any file permission issue.
Are you trying to install those fonts locally or use them by linking to Google's servers?
If you are trying to install them locally, you probably may not have the required permissions.
I've now downloaded those fonts and installed them on the wp server locally. Maybe wpastra can pick them up from there. Please check.
(When they said it still did not work, I responded as follows:)
1. Google fonts can be linked in web pages in such a way that the users of the web pages get the fonts
from google and display it on their browsers. If this has to work
properly, there are several things which have to be taken care of.
Example:
3. Another way in which fonts (any font) can be used is to just define the font, without embedding it - in this case, the font will fail to load if the same font is not present in the user's machine.
Without knowing what you are trying, whether the situation is 1, 2 or 3, I cannot help you further.
If you cannot tell me whether it is 1, 2 or 3, I would need access to a test page where the issue is seen.
(And they confirmed that it is working:)
The font issue is resolved, and we can choose the font type from the font family. We found out that we were missing one step for changing the font.
Sorry, perhaps I should have asked you for a small sample recording
before you recorded the whole thing. The audio quality is not great.
There is something wrong either with the settings or with your hardware
or both.
1. The same issue as last time,
the audio sounds as if it has been recorded over a telephone line or
something like that - the frequency response falls off after 5kHz, and
there is nothing beyond 8kHz -
You can compare that with this recording made with a Zoom H1 recorder,
where the frequencies up to 15 kHz and even a little beyond are clearly present.
2. Another bigger problem is that the recording has distortion due to getting clipped - the mic gain was "too hot".
(It turned out that the issues were due to recording using the laptop's built-in mic instead of using an Audiogram and a proper mic.)
As mentioned in September last year, Stellarium now has some features which enable creation of flybys. Perhaps Saturn is the most spectacular? Tried out creating two fulldome versions suitable for planetariums, one with Saturn visible at the apex of the dome, and another with Saturn near the front side of the dome, suitable for planetariums with unidirectional seating.
One of our servers which used to send a few notification emails to us via postfix stopped doing so - status=bounced (host gmail-smtp-in.l.google.com[142.251.2.27] said: 550-5.7.26 This mail has been blocked because the sender is unauthenticated. 550-5.7.26 Gmail requires all senders to authenticate with either SPF or DKIM.
By default, Kodi on LibreELEC running on Raspberry Pi 4 was detecting our Optoma ZK 507 projector and setting the output resolution to 4096x2160 @ 30 Hz. This was all right, but our 3840x2160 and 1920x1080 shows were showing a black bar at the bottom. Checking for solutions - not the minimise black bars global setting, but instead, setting the resolution from the comments on this thread,
I did not need to make a whitelist or blacklist in Kodi (blacklist is for devices, whitelist is for allowed video modes) - just setting the resolution at Settings --> System --> Display to 3180x2160 @ 30 Hz did the trick. Now, no black bars at the bottom of the FHD or UHD videos.
Another issue was, hiss in the sound output via a generic USB sound output, when amplified by our sound system. We set up a limiter and a gate with our old Behringer MDX 2200 compressor. Then, for further sweetening, I got a second-hand Lexicon Alpha sound device via bajaao.com for approximately one day's salary, and used that for sound output - no more hiss.
For recognizing the Lexicon Alpha, we had to plug it in after the Raspberry Pi booted into LibreELEC and Kodi - on every reboot. After recognizing the device, a one-time configuration change in Kodi --> Settings --> Audio settings --> Audio device , and set the output device to Lexicon Analog (not Lexicon SPDIF in this case).
Trying a direct connection for the Lexicon Alpha onto the RPi (as against a connection via a USB hub) also did not resolve the issue of not recognizing it on boot. So, we're plugging it in after every reboot (on the USB hub for convenience.)
That link also discusses some pros and cons - like no web lookups for huggingchat. So, I looked up how we can make our own chatbot, the full detailed step by step process, how much manual work is involved, and what we need to customize it.
(this seems to be the generic steps without customization)
So, link (2.) seems to indicate that we need to train it using a csv file with sample queries and responses. This is something which, for example, a large number of volunteers could generate. If each one generates questions and answers based on one Discourse, for example, 200 or 2000 people could generate quite a good data set.
We
could use AAAA records (IPV6) instead of A records (IPV4) to avoid
these charges, since Cloudflare would proxy to IPV4 and IPV6. In case we
need to SSH into the servers, and our ISP / router does not support
IPV6, then we might need something like this,
1. Go to your VPC settings and assign a /56 prefix. 2. Go to subnet settings within the VPC and assign a /64 to the subnet. 3. Go to the instance's network interface settings and enable IPv6 address assignment.
Logged in success with browser-rendered terminal on Edge browser on Windows 11.
Possible pain point - Note is that for each login (with a timeout of 24 hours as the default), we have to copy-paste the private key, in PEM format, into the authentication form. PEM format for the private key is a must. OpenSSH format doesn't work - Browser SSH Private Keys not working - Zero Trust / Access - Cloudflare Community.
I clicked on publish, and after going through the 4-5 steps verifying the scopes etc, I see the following message:
The Trust and Safety team has received your form. They will reach out to
you via your contact email if needed. The review process can take up to
4-6 weeks. Expect the first email from our Trust and Safety team within
3-5 days. Your last approved consent screen is still in use.
On one of our servers running Ubuntu 22.04, Wordpress reported upload size is set to 2 MB. Changed to 50M from 2M in the global php settings at /etc/php*/8.1/apache2/php.ini - that has fixed the issue.
We have used two different methods to deal with this - one solution using XOAuth, and one using App Passwords.
1. For some of our Moodle instances which used Google Workspace emails for outbound emails, we could use the in-built XOAuth support as explained here,
So I created two separate sets of credentials for one of the email ids, since the redirect url for each (one production and one development) server would be different as mentioned in the documentation linked from the tracker link above -
2. For our internal server using ssmtp as mentioned in this earlier post, this procedure (or the use of PHPMailer which supports XOAuth) would not be suitable, since it does not have a public-facing website url.
describes a python solution, but needs refresh, and needs a browser.
)
In passing, the email from google mentions that App Passwords are not going away - "Password-based access (with the exception of App Passwords) will no longer be supported" . We had earlier created App Passwords for use with We can create App Passwords if we enable 2FA. So, we enabled 2FA, created an App Password, and just replaced the password in our earlier ssmtp configuration file with the App Password (without spaces). And that works.
When a moodle server had its domain changed, the sendgrid account it used to send password reset emails was still using the old domain. When the "From" email id was changed, sendgrid gave an error,
550 The from address does not match a verified Sender Identity.
This moodle plugin "Dictionary" requires a one-time free registration to the Merriam-Webster api at https://dictionaryapi.com/ - they currently provide 1000 free lookups per day per key, and 2 free keys per registration. The plugin uses one key for dictionary, one for thesaurus.
Instead of commenting out, used the method mentioned lower down in the thread for Moodle 4.1 and above, by going to Site administration --> Server --> Support contact and changing the option to make it available only for logged-in users. Did the same thing for two of our production instances, too.
Paid tech support by email/telephone - Nettigritty.com
We have found that they usually respond within 3-4 hours.
But
a caveat - they would only extend support for those services for which
you have paid them. For example, if you have paid for web hosting, they
will support you only in web hosting related issues. They may not be
able to help you with open-ended questions or how-do-i-do-this-in-moodle
type of questions, though, of course, you could ask them if they will
do so.
As mentioned in my previous post, SheepIt render farm is a free render farm for Blender Cycles and Eevee engines, based on mutual co-operation - we connect our machines to the render farm to earn points, and can get our projects rendered in double-quick time in return.
Their FAQ page lists answers to most questions one may have about the service. I've uploaded an example of a project which includes the baked cache for particle systems, on Github, here. As mentioned in their FAQ, the steps to do this are:
1. Make all paths relative - in Blender, choose
File --> External Data --> Make All Paths Relative
2. If you need to render a baked fluid/particle system, zip the blender file and the cache directory, ensuring that the cache directory is also set as a relative path. In my case, I had set the cache directory to be inside the project directory to simplify matters. In the project above, this is located in the Physics tab of the Smoke Domain object as seen in the screenshot below. As you can see, the current directory is indicated by a leading double-slash - //.
3. One point to note is that choosing File --> Save As and saving the project to another directory, and then copying the cache directory there, does not work (as in Blender 3.2 on Windows.) This is probably because the file paths get mangled? Anyway, if you want to break up the project into multiple renders due to the bake files being too large, copy-paste the project outside Blender instead, don't use File --> Save As.
4. There is a script available on the Forums for splitting the project, in case you want to automate the process. Or, like I did, you can just choose to bake all, and then copy only the .vdb files in the data subfolder of the cache for a limited number of frames. For the zoomed out view, for frames after 275, this was something like 30-50 frames per project. And choose to render only those frames. Then create a fresh zip file for the next set of frames. And so on.
Here are some project statistics, for the zoomed in project render, which was taking around 3 minutes per frame on my machine:
Summary
Storage used: 971.8 MB
Cumulated time of render: 9h01m
Points spent: 5,767 points
Real duration of render: 1h23m
On reference per frame rendertime: 0m51s
On my machine, the 600 frames would have taken 30 hours to render, and even on the Studio Mac it would have taken 5-6 hours, but SheepIt finished it in less than 90 minutes due to the power of distributed processing.
So, nowadays I leave the sheepit client running whenever I switch on my machine, to accumulate points. A borrowed CPU from VV fitted with the GTX1050 Ti card which I purchased for my research also helps at the office. The 1050 Ti card is only 4% the speed of the standard GPU, but since most of the projects specify GPU, it usually gets some project to render, as against the i5-1235U laptop CPU, which often goes many minutes with "No jobs to render".
Edit: 17 Nov 2024 - interesting to see the bump in capability due to upgrading to SSD and running Linux on it, for the same machine - Power CPU and GPU went from 34% to 46% and 4% to 12% respectively -
And finally, here's a youtube video of the rendered output. This is a "Fulldome" video, meant for projection in a planetarium. (Since this is my first effort, the launch needs more work to look more realistic - currently the motion of the rocket looks a bit cartoonish.)
A quick note to myself - when importing image sequences into Blender VSE, we must ensure that the files are sorted by name in the file open dialog box (or in whatever order we want them to be) before selecting the relevant images, and then importing. Or else, some jumbled order will result in the imported sequence.
Its USP seems to be that it can save you from having to tap through many clicks on apps using routines which can be launched by the push-to-talk device. If it really delivers on the promises given in the talk in the link above, it might become the "next big thing."
As noted in the previous post, time-remapping to make the animation slower does not work well with particle systems, probably because baking is not done for fractional frames. A similar issue is discussed here.
A gotcha for me was that the keyframes for the rocket animation were not visible in the timeline, but some other keyframes, like the camera field of view animation which I had added, were visible.
So, dragging these keyframes to newer locations further along the timeline was not enough - I had to go to graph editor, and make sure the filters had "Show hidden" and then the keyframes (after 1000 frames) were visible.
More notes and links to tutorial video timestamps below,
under Performance, Memory, Tile Size. Only for memory's sake, so we need not bother about tiling for our renders.
On a Mac at Studio, renders a frame in around 26 seconds with CPU and 18 seconds with GPU. So maybe 5 hours to render 1000 frames, as against a couple of days on my i5 machine without GPU.
Sheep-It distributed free rendering - leaving the machine on and connected for around 4 hours earned 1600 points. There are some caveats to using baked particle systems as mentioned in their FAQ (Does SheepIt support fluids?), need to zip the folder and make sure it is less than 750 MB, but non-particle renders should be do-able. With the 1600 points as noted, my project was halfway up the queue, and started rendering after a 5 minute or so wait. Or maybe the wait was shorter, just needed to go to the Projects page to see an update. 100 frames done on 5 minutes, which on my machine would have taken 5 hours or so. But the "shephering" server might have been busy, so the last frame was shown as "in progress" for a long time - 10 minutes or so. But since we get an email when the project is done, actually no need to sweat.
The way to play local videos as a playlist seems to be, as per this forum thread,
right-click or C in the Files view to get the context menu for a video file, choose to queue it, the file view automatically jumps to the next file in the folder, right-click or C and choose to queue it and so on, and when all the required files are added, left arrow for the hidden menu, choose Go to Playlist, then choose Save - it will now ask you for a filename for the playlist.
In order to play this playlist, from the home screen, choose Videos (NOT Movies) choose Playlists highlight the desired playlist, C for context menu, and choose Play. Or press P to play after highlighting the desired playlist, after which you have to press TAB for full-screen video without the menu.
Most of the options like audio device are in the Kodi Settings (gear icon on top of home screen.) LibreElec settings for system name, keyboard layout, etc.
We have a Sony HD player for playback of 1080p files - works well for XVID-encoded files. When we had issues with the Firestick for playback, we tried out the Sony UBP-X700 4K Ultra HD Blu-Ray Player even though it was very much more expensive - Rs. 30k compared to the Firestick 4K's Rs. 5k. But unfortunately, it did not work out.
Problems with the Fire TV stick 4K were:
Using the Amazon Firestick 4K for playout is extremely flaky - since the
firestick keeps checking Amazon for updates and auto-updates, sometimes
the show stops playing.
Audio has to be taken from the projector's EP audio out, and has a lot of hiss.
Problems with the Sony were:
It was starting up in 4K 60 Hz which our projector did not support
It was changing modes when playing files with different resolutions and frame rates, causing delays and missed frames between files.
So we returned that, and are leaning towards playback using a Raspberry Pi 4.
The earlier models of R Pi, which we already have, do not support
4K. Raspberry Pi 5 has also come out now, which is slightly more
expensive, but which we don't necessarily need.
Estimated prices:
Raspberry Pi 4 - Rs. 6,500
Power supply - Rs. 1,000
HDMI mini to HDMI adapter - Rs. 664
The following items may also be required, which we might buy later if required:
Keyboard - Rs 300
Mouse - Rs. 300
HDMI mini to HDMI adapter - Rs. 664
(we may need 2 of these)
USB sound card, high quality - Rs. 2,000
Case with cooling for R Pi - Rs. 2,000
Mini SD card - Rs. 600
Edit: Raspberry Pi 4 running Kodi and a pre-owned external USB Lexicon Alpha audio interface is what we're using as of May 2024.
