Hello Thomas Dickey, please excuse the late reply.
Thomas Dickey wrote in <20190910092328.5aevtmukues22322@prl-debianold-64\ .jexium-island.net>: |On Mon, Sep 09, 2019 at 03:43:59PM +0200, Steffen Nurpmeso wrote: |> Thomas Dickey wrote in <20190908105750.qc3ut2e3rkrswfm4@prl-debianold-64\ |> .jexium-island.net>: |... |>|The interim labels (no upload-foo, etc) are to verify fixes for bug \ |>|reports. |>|I marked the "real release" steps with "*", above. Counting all of the |>|steps, I have a few hundred "foo" scripts (perhaps I should write \ |>|a script |>|to count them). |> |> That leaves me speechless. | |I've mentioned some of it here | |https://invisible-island.net/scripts/readme.html#build_framework I see. I collected some scripts over the years, too. I find that it helps to become more dumb, as over the years i forget how to do things, since the scripts put me on an island. How often did i want to do a thing just slightly tweaked, and had to look into the script which does the usual thing, and then wander through manuals for a long time just to put the minimal tweak into action. |https://invisible-island.net/personal/lint-tools.html Interesting read. Compilers and their current (mis-)directions, bugs etc. are a permanent source of annoyance and problems also for me. I even had to rewrite a complete library in a very short time in order to be compatible to g++ 3.4.2, which did not like unnamed unions. I mean, it is clear that perfection is impossible, and that bugs can always happen, but very often it is just plain lack of interest, not to talk about passion or faith. Looking at the crux IRC channel log i track (started using IRC in 2019!) i read 14:54 < ryuo> huh. good thing i never installed exim. 14:55 <@frinnst> all software is shit. just update and you will be fine That makes me said, i for one am out for something different. You name it in there regarding compilers and analysis, and i am really thankful for Coverity, too, and would wish the hundreds of megabytes that the compiler suites bring onto the boxes would go more in that analysis direction. Coverity found bugs for my software, from which i then could not understand why gcc nor clang did not find them. On the other hand i never used other automated tools for code inspection. I tried them shorty after Y2K i'd say, including memory debug libraries (was it electric fence or so), but it was terrible. Also i hate debuggers from the ground up. So I started instrumenting all my functions manually, on function enter and leave, to be used for not-yet-dead chirps (crash log), or timing graphs. More or less MT neutral especially the latter, but all just using a spinlock if available for global access. And i had written a really good memory layer i am still prowd of, with lots of statistics and upper and lower canaries, and memory accesses only if memory was verified to be accessible, and such things. Read only violations that did not catch, though. Therefore i even have to like that Google ASAN thing. |As I recall it, up until about 2005 I just did configure/make/install, |to spot-check things, but the packagers kept pointing out that I missed |things (either in builds or website maintenance), and I started adding |scripts to do more systematic builds, and assist in the website maintena\ |nce. |Starting in 2010, I switched to developing mostly in virtual machines, |which (though some rpm scripts were older) led to most of the package-sc\ |ripts. Oh, luckily i again have good hardware now and am able to build up a growing pool of VMs! That is really a tremendous improvement, having a local pool of myriads of different operating systems, at your fingertip! And at almost native speed. |Then in 2015, I started doing the routine export-to-git, which led to |doing more systematic labeling of snapshots. | |Still, with several steps, I occasionally overlook one (or find |that a script doesn't do what I meant). I bet it will never be perfect. Just when you think you now have a really good automatized environment, it will not take long until something catches your attention. Or your mind ticks and you suddenly see a different way that would do all the things even better. |> Yes, i have no idea of Windows at all. (Except having had the |> Cygwin source code to have a look sometimes, to not end in total |> dead ends shall i ever port to it.) | |I've a few machines with cygwin, mingw and msys2, and use the no-cost |visual studio versions, and collect logfiles. Aside from the X11 stuff |such as xterm, other programs "should" build/work with Windows console. | |Not everything gets built at once, and some are old, |but I've a large number of those for reference (another |opportunity for counting/reporting, here). That is great. I unfortunately not, yet. Hopefully i will make that happen later. I did not even try out FreeDOS, even at its jubilee, no time yet. |I made a to-do item for the script |(since a nice table would improve the webpage), |for later, sometime. | |>|If you have something that works reasonably well on one of the BSDs, |>|making it more portable (adding configure checks, command-line options, |>|etc.) is fairly routine. |> |> That is BSD socket code + read(2) + write(2). |> I will throw an eye as soon as possible, and report back then, ok? | |sounds good I hope i find time tomorrow! It is just, i just cannot do something if i have another thing in my mind, it will not turn out the way i want it, anyhow. (Especially since i have no glue of and to the coding side of lynx.) Ciao, --steffen | |Der Kragenbaer, The moon bear, |der holt sich munter he cheerfully and one by one |einen nach dem anderen runter wa.ks himself off |(By Robert Gernhardt) _______________________________________________ Lynx-dev mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/lynx-dev
