The forum is dead for now... and to be honest it has been more work than desired; php updates, spam, software updates... hasn't really saved time as planned. But that isn't the point for today. I just wanted to make some random rants about my current work. I was going to use the forum, but... anyway so for now I'll go low tech and just add pages to my site.
GigAlarm Alpha is now on the Beta page. This has been undergoing a lot of testing, and there will be much more. This will probably be one of the last Beta's to be finalised. Next up, Chop. We are hoping to up the release cadence soon...
IPScour Beta / alpha is now on the Beta page. This one is a bit weird as it is a long term beta that will not be released at the end of this whole 1.9 process anyway. So it is really just another Beta. Like last time another is on the way, GigAlarm. It is still undergoing testing but will hopefully be the next Alpha...
Break v1.9 Alpha is now on the Beta page. It has been used locally for the last few weeks so it is OK for a public Alpha now. IPScour is probably next.
BrickCol v1.9 Alpha is now on the Beta page. Break is on the way too but it is a lot more complicated and needs more testing.... even for an Alpha release.
Yeah. As I said LEGO kept me busy last month. BrickBuilt 2022. Over now. Aftermath is still lingering, and I made a new Windows 11 computer last week. But there's no excuses left this year. Pretty sure....
Ummmm busy with LEGO this month. Yeah. Seriously.
So yeah, only got XorIt v1.9 on the Beta page. Still hoping to get another done this month but time is running out.
Versus v1.9 is now on the Beta page. More ranting later.
Never got around to saying this, but this whole 1.9 release thing got delayed as after the code was all brought together the next (unplanned) job was other build chain related updates. Which were more involved than planned. But they are done! So now it is the next stage; betas! Hopefully there will be some new content soon!
That last post was supposed to be quickly followed by a more useful update. Didn't happen though as there was too much going on over the silly season... and it is gonna be busy for the next month too. This Slog started in April 2021. Let us see if there can be something more solid happening before then...
Last post of the year? Time will tell...
Darn. Another month with only one update. So yeah. Still at it. Most of the MASM32 programs have been fixed so they assemble, but we are waiting for all to be done. Then we will be re-releasing all of them. This was the orginal 1.9 plan.... mentioned somewhere in the forum... In short, test assemble everything, archive the code, and then bring out 2.0 for the more popular ones. In the end only iTimeSync got that release, with XorIt being a long term beta... Not sure what will end up happening now. Might end up reversioning the C++ ones as 3.0...
So after much soul searching we've decided to give our assembly releases one more... release. Why? Well it's been clear that some programs will not get converted for some time. Yet some need a few tweaks due to the world moving on. (Including some that we use a lot!) Also, during the porting of TwinSeek we were having trouble finding the latest version... and... we wanted the old version to compare to but it was hard to debug. At the core of this issue is over the last ten years we're been tweaking programs on various machines and (shock horror) we were not using any source control. Why? MASM32 really is designed to be installed in a specific location and as a result it was only moved when our primary machine changed. We have now though retweaked the way the projects are built so we can now use git and use multiple machines without multiple confusion. So, in short this is one of the reason for the lack of progress. Hopefully next month we'll have some more changes around here.
Huh... where did August go... well I am sure I will get loads done in September... oh...
So as I said, things might take longer depending on priorities and life and stuff. That is the case, for better or worse. No ETAs. Will try to get back on the case ASAP. Did fix some annoying web issues today, caused by the clean up of our domains and hosting. But for now, this is the news.
So still on Egg. That tool, a directory change monitor, was done... eventually it might be added to TwinSeek, but making it nice is often the trick. It is easy to write something when all it has to do is match your thought process. Making software that everyone, or at least most, can understand takes a lot more work. It could be better to split it off. (Right now it is a TwinSeek fork.) In any case it will not be added to the current Beta. No current ETA on the Egg Beta. It is generally is easier to do a port from the middle and then write moving out as then the hierarchy is clear. But that is probably just an opinion. At this current point in time it is unclear how much is done and is not... sure it could be roughly calculated but it is not productive use of resources, and since often rewites/restructures are needed it is not a simple as counting lines done and lines not done. It will be done when it is done. Hopefully this week. Might be longer depending on priorities and life and stuff. By then it will probably be time to finish off iTimeSync. It is actually quite tiring porting assembly to c++. Some aspects, like stack and register handling, have no direct translation and the intent needs to be determined, which take time. But on that subject it is time for sleep here. Zzzzzzzzz.....
So the QWave beta is done. While this may have seemed to be an odd priority we wanted to do this mostly as it was important to find the size baseline. MASM32 programs can be really small; QWave was only 6kB... yet it seems the smallest stub program in VS C++ is about 12kB... and QWave 2.0 is currently 15kB. However QWave 2.0, unlike the current one, actually still works. Not sure what is the issue, probably Unicode is default now and that was never supported. Might look into it at a latter date. In any case the C++ versions will be at least slightly bigger, which is what has been seen so far. iTimeSync for example is about 50% bigger. Do note though that we are not compiling for size; optimised C often has unrolled loops for example. This is OK in reality. Maybe we should ret-con ACAP to be As Compact As Practical? Hmmmm.... Anyway, so what is next? Well there is some other research to do that porting QWave enabled, but probably Egg as it still is quite popular, although the temptation is there to do easy ports like Slam, Volt and Stims. Probably not though. Could be smarter to alternate between a big port and a small one. QWave 2.0 highlighted that it can be done quite fast. In fact more time was spent on re-arranging code for minimum size, writing tests, and rebuilding/testing other programs affected by these changes. Might do some more work on iTimeSync, but there is no intention to finish it for at least a few weeks because, as expected, working on different programs highlighted improvements for all. But some of the missing features like server scanning might be finished. Also, still have not written that tool we wanted.... and this rant is long enough. Back to work!
The 99% complete TwinSeek has been updated on the Beta page. A bit later than planned, but not too bad. There were some issues porting some functionality and as a result it became easier to bring forward some planned unit tests. It is funny with trying to finish software. You have lists of issues and then you suddenly realise that it has all come together and everything seems done. Still loads of testing to do, and the release will be after iTimeSync so it will be at least another two weeks. So, the next progam to appear on the site is planned to be QWave, but there is a tool that is wanted for local reasons so that may happen first. Time will tell. Off for now...
iTimeSync has been added to the Beta page. There are some known missing features so it is being declared as a Alpha. There should be a final Beta in the next few weeks that will move to a release a week later. Or so the plan goes. Right now the plan is to move TwinSeek to a final Beta. That will be released first; it has been very well tested and in use for years. There were just some UI issues. Hopefully that will be in under a week. It does seem that this process will take longer than envisioned. While MASM32 is structually very similar to C the delay is (1) we are going for a C++ -ish port so while programs were done in a object-orientated fashion it was in assembly so it was bendable and sometimes broken for various reasons. But the bigger reason is (2) very similar is not the same, and as a result it becomes easier often to just retype the entire program while referring to the MASM32 original. This takes a while. Sigh. Final bit of odd news is QWave will be ported soon. Enough ranting. More work to do.
Still working on iTimeSync. Probably about 90% ready for a Beta, but that'll probably be a few more days at the earliest. Other programs should be faster; a lot of the work on iTimeSync was library update and design. We had libraries of sorts from early days, but we allowed forking within programs to a degree in the interest of efficiency. Less inclined to do that now as it became a big reason why minor work became harder than it should have been. This is why we want to wait until we have a few active Betas before we release any.
So yeah, I'm in the middle of a complete rebuild of everything in C/C++. Probably. Some programs may be missed, but the goal is to hold back any release until I've got a few ready. TwinSeek 2.0 has been in active use for years, Ute is in progress, and iTimeSync is aproaching a Beta. There's another unfinished program as well but that'll be delayed until the rebuilds. More later. Still gotta add other links to make the page work.
| 
For software As Compact As Possible
