Posts tagged with "Coding"

Now that BlogMore is so obviously feature-complete and there's nothing else I could possibly need to add to it... here's v2.32.0 with a quick new feature.

The work to add the relatedness of posts got me wondering about other things that could be measured about the content of posts, and a bit of reading and looking for ideas resulted in my (re)discovering the Gunning fog index. That got me thinking it might be interesting to see what the highest and lowest values were for my blog, and perhaps what the average is, perhaps even get a feel for the distribution.

So now BlogMore has a with_gfi configuration setting (and a matching command line switch). When enabled, it adds the score for a post next to the reading time (if that's enabled, if not it's in the same location as the reading time would be). It also adds a load of stats to the stats page.

As well as this change, I've also tweaked the stats page so it's easier to get a link to a particular section. As with headings in posts, when you hover over a section heading you'll get a little anchor you can click on that will result in you visiting that particular heading. That should make it easier to highlight a particular statistic, if you need to.

A quick update to BlogMore, with a small accessibility improvement, and also a whole new set of data for the graph.

The accessibility update follows on from a change published yesterday. In that release I did a little bit of work to disambiguate links to categories and tags that had the same link text; this time I've given the same aria-label treatment to the post dates where they link to the year, month, and day archives. While we're never really going to get ambiguous year links (unless there are also links in the body of a post that are four digit numbers), it's highly likely we're going to get the occasional month and day that are ambiguous.

So now those links make it clear, with aria-label, what's being linked to.

The other main change in v2.31.0 comes from a thought I had last night about the new "related posts" feature I added yesterday. Currently the graph shows the relationships the posts have based around their common categories and tags:

But now I've got this code that can work out how posts might be related based on their text content. What might that look like? It might be cool to be able to switch the graph view between the two sets of data, if the blog owner is building with related posts turned on.

While I'm not sure it really tells me anything, I like that I have yet another method of (re)discovering posts.

BlogMore has been bumped to v2.30.0. This release is pretty heavy on new features, but it does also include one small accessibility tweak too. While looking through some of the neutral feedback from PageSpeed Insights I noticed it mentioned that in some cases I had a category and a tag linked on the same page, where the text of the link was the same. That's pretty common on my blog. For many of the categories (especially things like AI, Coding, Emacs, Python, etc.) I'll also have a corresponding tag. The idea is that categories are essentially sub-blogs within the blog, whereas tags characterise a post.

Given that the same text for different purposes doesn't give much context from an accessibility point of view I've added appropriate aria-label attributes to differentiate these links.

Now for the new features.

The first is another "discover other posts" type feature, that might encourage a reader to venture further into a blog. While BlogMore does support backlinks (as was added back in v2.16.0), I have been thinking that a "related posts" feature would be interesting to add. So now it's added. This is enabled with the with_related configuration option (or the --with-related command line switch) and provides control over how many related posts are shown; the default is three.

Because calculating the best related posts, for every post, will cost in time and memory, some control over how "accurate" the hits are is given. You're also not stuck with "Related Posts" as the title either.

The rest of the new features are more admin-based and are all commands on the BlogMore command line. The first is the drafts command which simply lists the filename of all the posts that are in a draft state.

The second new command is links dump. This is a utility command for dumping a CSV list of all the external links that can be found in your posts, along with the filename of the post the link was found in. This could be useful for all sorts of things; for example writing your own external link statistics tool, or perhaps writing your own external link checker.

Talking of external link checkers: I've also added links check. This is a bit experimental, but is intended to be a simple checker of all the external links, seeing if they're still out there. By default, when run, it'll check every link it finds and, if there's a problem with it, it'll report the issue. There's a --verbose mode as well if you want feedback on all the links that are working.

It seems like every time I think I'm done adding features, something else is either suggested or pops into my head. I feel like I'm near the end of adding stuff now and should be getting back to refining the code.

I'm fairly sure I've said that before.

While it did appear something had changed for the better yesterday evening, things looked different again this morning. Before getting to work, I did a couple of quick changes to BlogMore; this time Antigravity seemed to use up a bit more quota in the same amount of time as I was working last night, but this time the reset window is much bigger.

I'm sure there's a method to this, but at the moment I'm really unclear what it is.

I guess I could ask it...

Yeah, let's do that.

Can you explain your quota cooldown time for me? Sometimes it seems to count down from 1 hour, others 5. Am I not reading this correctly?

You are reading the quota cooldown correctly! The reason the timer fluctuates between 1 hour, 5 hours, or other times is due to how the rolling usage window and compute-based limits work.

1. Rolling 5-Hour Window

Your usage limit is evaluated on a rolling window (typically 5 hours) rather than resetting at a fixed time of day.

• If you see a 5-hour cooldown: You likely used a large portion of your compute capacity very recently. You have to wait for that recent window to clear.
• If you see a 1-hour cooldown: You are close to the edge of your rolling window. A batch of high-compute requests you made 4 hours ago is about to fall out of the 5-hour window, which will free up capacity for you in 1 hour.

2. Compute-Based Usage (Not Message Count)

Your quota is consumed by compute rather than a flat number of messages. Compute is affected by:

• Chat Length: Every time you send a new message, the model re-processes the entire chat history. Longer conversations consume your quota much faster.
• Task Complexity: Tasks like code generation, deep thinking, or agentic workflows require more processing power and consume more quota.

Tips to Manage Your Quota:

• Start New Chats: For new topics or tasks, start a fresh conversation. This prevents the model from having to re-read long histories and saves significant compute.
• Use Clear, Targeted Prompts: Try to combine instructions rather than sending multiple short follow-up messages.

So, I don't think I'm any wiser after that. The explanation for the 1-hour countdown I saw last night has no connection to anything I did yesterday; I wasn't using Antigravity 4 hours earlier when I checked the quota near the end of that session. Guess I need to go and find and RTFM. Just like we used to in the good old days.

Something has changed with Antigravity since I first tried it out the other day. While looking at adding a related posts feature to BlogMore, I thought I'd give it another try out (having gone back to using Gemini CLI while I still could).

That first (and last) time I tried it, while on whatever model it decided for me out of the box, it chewed through most of the quota, with a 5 hour reset, in very little time at all. It was obvious that I'd never get anything of significance done in a good session.

This evening has been quite different. It wrote a very comprehensive change, doing quite a lot of work, and left me with a lot of quota and a short reset time once done.

A bit more testing and tweaking of the documentation followed, with me setting it off on a couple of bug hunts (which it found and fixed). By the time I was happy to call it an evening on this round of modifications, it had reset and I was green across the board again.

Now this I can work with!

I don't really know what's changed¹, or why. I think I saw something the other day about quotas being tripled, but this seems even more generous (at least in terms of the reset window). I guess I'm going to have to go digging to see if I can find what the story is.

I'm not getting my hopes up -- what can be given can be taken away at any moment (which is, of course, the ongoing theme of what I'm documenting here) -- but this does soften the landing somewhat.

A quick update to BlogMore that stems from a couple of accessibility concerns raised when using PageSpeed Insights. While the accessibility score on pages generated by BlogMore is pretty excellent (96/100), a couple of things were highlighted that seemed worth tackling in some way.

The first was that the title bar of code blocks is rather low in contrast, especially the name of the language. This was a deliberate choice on my part; when I'd prompted Copilot to do this work some time back I'd expressly asked it to make the text "subtle". While I could have undone that, I do like how it looks and would like to retain it if possible. So, as an experiment, I've used prefers-contrast as a signal that I shouldn't do that. So now there's a bit of extra CSS under prefers-contrast: more which just uses the normal text colour for those headers.

I imagine this won't boost the score on PageSpeed Insights -- presumably it won't simulate that being used -- but I would hope it's an improvement for those who do need the extra contrast.

The other issue that was flagged up was the lack of an underline on the email address link in the comment-via-email invitation box (the optional feature I added back in v2.17.0). That one was an easy fix. There was no reason to avoid an underline on the link, so I added one.

I've just released BlogMore v2.28.0. This release has some small improvements to the JSON-LD structured data that was added in the last release, and also adds support for actually showing author names on posts.

On the latter point first: I only ever really created BlogMore for myself, thinking that perhaps some other folk might use it at some point. All along, though, I had it in mind that it would only ever be used to create a site where there was a single author. Despite this, though, I'd added support for setting the author per-post, and this was reflected in the RSS and Atom feeds.

But I'd never added support for showing the author in posts.

So this release adds that feature. I've tried to add full control, with settings that let you turn on/off showing the author blog-wide, as well as providing control per-post. Also, as well as setting the default author for a blog, and being able to set the author for a specific post, you can do the same for an author's URL (setting it blog-wide and per post).

Obviously, when the author is shown, if a URL can be worked out (the one local to the post is chosen first, then the blog-wide default if one isn't set for the post), the author's name links to their URL.

The JSON-LD changes are a couple of small improvements to the content. The author data adds a url property (following the same rules mentioned above) and the image property for a post will fall back to the site logo (if one is set) when there is no cover.

Much like the last two releases of BlogMore, this is another that has ended up being on the theme of improving or adorning the generated HTML.

One change in the last release resulted in another HTML validator warning, and so that's cleaned up here (the removal of the h2 elements from the sidebar meant it no longer made sense for it to be a section, so I've turned it into a div).

On top of that, I've also decided to dip my toe into adding more "microformat" type things to the generated code. This release adds things like JSON-LD structured data and Microformats2 semantic markup, where appropriate. I've also updated all of the "socials" links that appear in the sidebar to ensure they're marked up as rel="me".

Given that this is a bit of an experiment, expect to see some tweaks and changes as I roll this out on this blog and then check and test the result. This is a useful learning exercise for me.

The previous release of BlogMore had some work that improved the HTML, ensuring that the HTML validator was happy with the generated code. Yesterday evening I ran it over more of the pages, and found a couple more things that made sense to address.

So v2.26.0 takes this a little further and tries to clean more things up. Changes include:

• The comment email invitation box is now created with a div rather than a section, resolving a validation error about there being no h2 or similar inside the section (because we don't need any kind of heading in there).
• Cleaned up an error relating to the misuse of an aria-label in the graph page.
• Cleaned up an error relating to the misuse of an aria-label in the stats page.
• Removed the h2 elements from the sidebar, making them into divs with the same style. This leaves headings as something that will only appear in the main body of any page or post.
• Added some heading-demoting to the rendering of posts so that the heading structure of any given post is retained when it's part of any of the archive-style pages (date archives, category archives, tag archives, etc).

While not all of the above were being reported as validation errors, all of them should result in HTML that better fits what I'd want in the first place.

As I've written about a few times in the last week or so, the journey with AI-based coding tools has hit an interesting time when it comes to prices, quotas, usage, availability and all that. Having come into all of this via a place where it was a flat fee, and where I didn't really need to think about input tokens and output tokens and so on, I'm pretty ignorant of what that all means in terms of scale. If I'm looking at a new tool and I see prices and/or quotas for in/out tokens, it means nothing to me. I can't relate to it. I've never had to care about it.

While using Gemini CLI to quickly make a change to BlogMore this morning, I was reminded that at the end of a session it does tell me this:

Seeing that got me thinking: is there a way to get the total usage for all of my sessions, or at least the sessions that have still been retained (I'm guessing they expire after a wee while)? After a little bit of searching I found ccusage. That looked exactly like the sort of thing I was after.

Now, this is only going to be good for Gemini (it says it supports Copilot too, but it seems to be failing to find any Copilot sessions), but it should give me a feel for what my token usage looks like.

I work on BlogMore on two different machines: the MacBook Air and also a Mac Mini I have in my office. Here's all of the available token usage data I can get out of the Air:

And also the same for the Mac Mini (which gets used less frequently for this sort of thing):

In both cases I've removed a couple of columns to make the tables fit better. The first was the model name (varying between gemini-3-flash-preview and gemini-3.1-pro-preview), the second was Cache Create (which was always 0 all the way down).

From what I can see, it would appear that these two tables do cover my increasing use of Gemini CLI for doing work on BlogMore (the first intensive use being back around the 5th of this month, if I recall correctly). So this would seem to be a reasonably informative way to view things.

All of which is to say, over a roughly three week period, while getting things done, I've used getting on for 20,000,000 input tokens, and around 600,000 output tokens (presumably I do also need to be keeping the 114,300,000 cache read tokens in mind too). With this in mind I might now be able to make more sense of the pricing I see for various tools.